Qwen3.8 Max
开始使用
API 接入
Cloudflare AI Gateway API文档
目录部署标识：alibaba/qwen3.8-max。具体区域、权限与价格以服务商为准。
Alibaba API文档
目录部署标识：qwen3.8-max。具体区域、权限与价格以服务商为准。
Vercel AI Gateway API文档
目录部署标识：alibaba/qwen3.8-max。具体区域、权限与价格以服务商为准。
AI/ML API API文档
目录部署标识：alibaba/qwen3.8-max。具体区域、权限与价格以服务商为准。
LLM Gateway API文档
目录部署标识：alibaba/qwen3.8-max。具体区域、权限与价格以服务商为准。
公开成绩与计算依据
综合参考分根据已公开的评测资料整理。查看发布日期依据
综合参考分 55.5 · 11项基准
推理求解 62.1
GPQA Diamond · 92.6
huggingface.co · 官方发布比较 · Diamond
Qwen3.8-Max · Official published comparison; GPQA Diamond; exact column Qwen3.8-Max. Footnotes and all competing runs retained in source.
Terminal-Bench-Science · 12.86%
Vals AI · 独立评测 · 0.1.0
Qwen3.8 Max / Mini-SWE-agent · Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.
工具与任务执行 58.4
APEX-Agents · 34.0
腾讯 Hy4 官方发布比较表 · 官方发布比较 · current published task suite
腾讯官方比较表 · Qwen3.8 Max · APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM
Terminal-Bench · 34.34%
Vals AI · 独立评测 · 4.0
Qwen3.8 Max / Mini-SWE-agent · Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.
AutomationBench · 27.3
huggingface.co · 官方发布比较 · current published task suite
Qwen3.8-Max · Published page qwen38-weights; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings
AA-Briefcase · 0.49595959595959593 比例
Artificial Analysis · 独立评测 · 1.1
Qwen3.8 Max · aa-briefcase-rubric-reviewed-method-20261029
编程 25.0
DeepSWE · 56.6
qwen.ai · 官方发布比较 · 1.1
Qwen3.8-Max · Published page qwen38-max; row DeepSWE 1.1; preserve page footnotes and task-specific settings
语言与信息处理 67.7
IFBench · 82.8
qwen.ai · 官方发布比较 · current published task suite
Qwen3.8-Max · Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings
SimpleQA Verified · 0.458 比例
Epoch AI · 独立评测 · run 1.2.0
qwen3.8-max_xhigh · epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0 · 2026-08-27公开
多模态理解 69.0
Humanity Sixth Sense · 33 %
Scale AI · 独立评测 · current published task suite
Qwen3.8-Max (max) · scale-hss-reviewed-method-20261029
MMMU-Pro · 82.3
qwen.ai · 官方发布比较 · Standard 10
Qwen3.8-Max · Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings
资料来源
- 版本日期来源 · 版本资料 · 2026-10-29查阅
- huggingface.co · official_release_comparison · 2026-10-29查阅
- Vals AI · independent_evaluation · 2026-10-29查阅
- 腾讯 Hy4 官方发布比较表 · official_release · 2026-10-29查阅
- Vals AI · independent_evaluation · 2026-10-29查阅
- Artificial Analysis · independent_evaluation · 2026-10-29查阅
- qwen.ai · official_release_comparison · 2026-10-29查阅
- Epoch AI · independent_evaluation · 2026-10-29查阅
- Scale AI · independent_evaluation · 2026-10-29查阅