List your company

End-to-end Tool-use on ToolBench I3 v1

52.19SoPRToolWeaver
9.279620.419831.5642.7002Jan 29, 2026

Evaluation results

MethodLinks
ToolWeaverSetting=Direct Generation2026.0152.1959.02
ToolWeaver2026.0152.1959.02
ToolLlama*Setting=Retrieval2026.0151.3731.15
ToolGen*2026.0147.5434.42
ToolGenSetting=Direct Generation2026.0136.3449.18
ToolGen2026.0136.3449.18
ToolLlama-2*Setting=Retrieval2026.0134.721.31
GPT-3.5*Setting=Retrieval2026.0134.4337.7
GPT-4o-miniSetting=Retrieval (Re.)2026.0124.86—
GPT-4o-miniSetting=Retrieval2026.0124.8644.26
ToolLlama-2Setting=Retrieval (Re.)2026.0110.9324.59
ToolLlama-2Setting=Retrieval2026.0110.9324.59