Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Tool-use Task Completion on ACEBench agent subset Multi-Turn
Loading...
84.4
Success Rate
ReAct + EvoSOP
45.192
55.371
65.55
75.729
Jul 8, 2026
Success Rate
Updated 17d ago
Evaluation Results
Method
Method
Links
Success Rate
ReAct + EvoSOP
Backbone=Qwen-Max
2026.07
84.4
DFSDT + EvoSOP
Backbone=Gemini-3-FP
2026.07
83.3
ReAct + EvoSOP
Backbone=Gemini-3-FP
2026.07
79.4
ReAct + EvoSOP
Backbone=GPT-4o
2026.07
79.4
DFSDT + DRAFT
Backbone=Qwen-Max
2026.07
78.3
DFSDT + EvoSOP
Backbone=Qwen-Max
2026.07
77.9
DFSDT + EvoSOP
Backbone=GPT-4o
2026.07
74.4
ReAct
Backbone=GPT-4o
2026.07
72.2
DFSDT + DRAFT
Backbone=Gemini-3-FP
2026.07
71.1
DFSDT
Backbone=GPT-4o
2026.07
69.4
DFSDT + DRAFT
Backbone=GPT-4o
2026.07
66.1
ReAct + ASI
Backbone=Qwen-Max
2026.07
65.6
ReAct + ASI
Backbone=GPT-4o
2026.07
57.2
ReAct + ASI
Backbone=Gemini-3-FP
2026.07
46.7
Feedback
Search any
task
Search any
task