Agnes 2.5 Pro Beta
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 66.9 · 4项基准
推理求解 80.5
GPQA Diamond · 0.905050505050505 比例
Artificial Analysis · 独立评测 · Diamond
Agnes 2.5 Pro Beta · Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond
工具与任务执行 67.1
tau3-Banking · 0.356701030927835 比例
Artificial Analysis · 独立评测 · 1.0.1
Agnes 2.5 Pro Beta · Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking
GDPval-AA · 1335.33 分
Artificial Analysis · 独立评测 · 2.1
Agnes 2.5 Pro Beta · Artificial Analysis current published protocol; gdpval-aa-normalized
多模态理解 32.4
MMMU-Pro · 0.754335260115607 比例
Artificial Analysis · 独立评测 · Standard 10
Agnes 2.5 Pro Beta · Artificial Analysis current published protocol; mmmu-pro