List your company

Tool-use API Generalization on ToolBench G2

78.2Pass RateEVOTOOL
41.851.2560.770.15Mar 5, 2026

Evaluation results

MethodLinks
EVOTOOLBackbone=GPT-4o2026.0378.2
DRAFTBackbone=GPT-4o2026.0376.8
EASYTOOLBackbone=GPT-4o2026.0375.1
EvoPromptBackbone=GPT-4o2026.0369.8
ANYTOOLBackbone=GPT-4o2026.0368.8
OPROBackbone=GPT-4o2026.0367.3
PromptBreederBackbone=GPT-4o2026.0366.5
EVOTOOLBackbone=Qwen3-8B2026.0366.5
DRAFTBackbone=Qwen3-8B2026.0365.3
ReActBackbone=GPT-4o2026.0364.7
EASYTOOLBackbone=Qwen3-8B2026.0361.8
EvoPromptBackbone=Qwen3-8B2026.0359.3
Plan-and-SolveBackbone=GPT-4o2026.0359.2
AdaPlannerBackbone=GPT-4o2026.0358.5
ANYTOOLBackbone=Qwen3-8B2026.0358.5
OPROBackbone=Qwen3-8B2026.0357.2
PromptBreederBackbone=Qwen3-8B2026.0356.5
ReActBackbone=Qwen3-8B2026.0355
CoTBackbone=GPT-4o2026.0350.8
Plan-and-SolveBackbone=Qwen3-8B2026.0350.3
AdaPlannerBackbone=Qwen3-8B2026.0349.7
CoTBackbone=Qwen3-8B2026.0343.2