List your company

Function Calling on ToolBench Average

75.95Pass RateParaTool
25.35438.489551.62564.7605Jan 26, 2026Feb 15, 2026Mar 7, 2026Mar 28, 2026Apr 17, 2026May 7, 2026May 28, 2026

Evaluation results

MethodLinks
ParaToolBackbone=Qwen2.5-7B-Instruct2026.0575.9580.75
ParaToolBackbone=Llama3.1-8B-Instruct2026.0569.5771.82
ToolLLaMABackbone=Qwen2.5-7B-Instruct2026.0568.3170.18
Parameter+ReinvokeBackbone=Qwen2.5-7B-Instruct2026.0568.2871.28
Context+Docs & ExamplesBackbone=Qwen2.5-7B-Instruct2026.0566.9669.08
Context+Docs & ExamplesBackbone=Llama3.1-8B-Instruct2026.0565.6763.58
ToolLLaMABackbone=Llama3.1-8B-Instruct2026.0565.4165.5
Parameter+ReinvokeBackbone=Llama3.1-8B-Instruct2026.0563.7865.13
Claude-3.5-Sonnet + TAFCFramework=TAFC, Temperature=0.1, Independent Runs=32026.0160.372.2
Parameter+ToolRetrieverBackbone=Qwen2.5-7B-Instruct2026.0560.2662.76
Parameter+ToolRetrieverBackbone=Llama3.1-8B-Instruct2026.0560.0560.6
GPT-4o + TAFCFramework=TAFC, Temperature=0.1, Independent Runs=32026.0159.271.4
Claude-3.5-SonnetFramework=Standard FC, Temperature=0.1, Independent Runs=32026.0158.470.5
Context+DocsBackbone=Qwen2.5-7B-Instruct2026.0558.06—
GPT-4oFramework=Standard FC, Temperature=0.1, Independent Runs=32026.0157.569.4
Parameter+EmbsimBackbone=Qwen2.5-7B-Instruct2026.0557.0461.91
Parameter+EmbsimBackbone=Llama3.1-8B-Instruct2026.0556.2157.59
Context+DocsBackbone=Llama3.1-8B-Instruct2026.0556.13—
Qwen2.5-72B + TAFCFramework=TAFC, Temperature=0.1, Independent Runs=32026.014957.1
Qwen2.5-72BFramework=Standard FC, Temperature=0.1, Independent Runs=32026.0146.955
Llama-3.1-70B + TAFCFramework=TAFC, Temperature=0.1, Independent Runs=32026.0146.454.4
Llama-3.1-70BFramework=Standard FC, Temperature=0.1, Independent Runs=32026.0144.252
Qwen2.5-32B + TAFCFramework=TAFC, Temperature=0.1, Independent Runs=32026.0142.750.8
Qwen2.5-32BFramework=Standard FC, Temperature=0.1, Independent Runs=32026.0140.548.3
Global ParameterizationBackbone=Qwen2.5-7B-Instruct2026.0539.3338.19
Global ParameterizationBackbone=Llama3.1-8B-Instruct2026.0535.232.56
Qwen2.5-7B + TAFCFramework=TAFC, Temperature=0.1, Independent Runs=32026.0131.238
Llama-3.1-8B + TAFCFramework=TAFC, Temperature=0.1, Independent Runs=32026.0129.736.3
Qwen2.5-7BFramework=Standard FC, Temperature=0.1, Independent Runs=32026.0128.734.9
Llama-3.1-8BFramework=Standard FC, Temperature=0.1, Independent Runs=32026.0127.333.4