Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Agent Capability Evaluation on ACEBench Agent

95Multi-Step Reasoning Score

GPT-4.1

32.648.86581.2Jan 9, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.01
956077.5
2026.01
8563.3374.17
8573.3379.17
2026.01
856072.5
2026.01
83.336071.67
2026.01
8063.3371.67
2026.01
8061.1170.55
2026.01
73.336066.67
2026.01
66.6733.3350
2026.01
63.3356.6760
2026.01
58.3352.2255.28
2026.01
5532.2243.61
2026.01
3528.8931.95