List your company

End-to-end Tool-use on ToolBench v1 (I2)

52.2SoPRToolGen*
23.589631.017338.44545.8727Jan 29, 2026

Evaluation results

MethodLinks
ToolGen*2026.0152.262.26
ToolLlama*Setting=Retrieval2026.0149.9661.32
ToolLlama-2*Setting=Retrieval2026.0149.2153.77
ToolGenSetting=Direct Generation2026.0145.1342.45
ToolGen2026.0145.1342.45
ToolWeaverSetting=Direct Generation2026.0144.0348.11
ToolWeaver2026.0144.0348.11
GPT-3.5*Setting=Retrieval2026.0141.1953.77
GPT-4o-miniSetting=Retrieval (Re.)2026.0140.41—
GPT-4o-miniSetting=Retrieval2026.0140.4152.83
ToolLlama-2Setting=Retrieval (Re.)2026.0124.6930.19
ToolLlama-2Setting=Retrieval2026.0124.6930.19