List your company

Tool-use Inference on API-Bank

5.8Match Rate (#MAT)TOOLSPEC
0.8082.1043.44.696Apr 15, 2026

Evaluation results

MethodLinks
TOOLSPECModel=LLaMA-3.1-8B-Instruct2026.045.8125.594.454.19
TOOLSPECModel=LLaMA-3.2-3B-Instruct2026.045.2133.34.093.66
TOOLSPECModel=Qwen2.5-7B-Instruct2026.044.9129.43.953.67
TOOLSPECModel=Qwen2.5-14B-Instruct2026.044.4375.273.863.48
Eagle-2*Model=Qwen2.5-14B-Instruct2026.043.8746.82.42.12
SAMDModel=Qwen2.5-7B-Instruct2026.043.683.962.562.22
TRModel=LLaMA-3.1-8B-Instruct2026.043.3471.12.522.45
Eagle-3*Model=LLaMA-3.1-8B-Instruct2026.043.3447.241.671.56
SAMDModel=LLaMA-3.1-8B-Instruct2026.043.269.682.472.41
TRModel=LLaMA-3.2-3B-Instruct2026.043.1175.52.322.32
TRModel=Qwen2.5-7B-Instruct2026.043.0978.952.412.28
TRModel=Qwen2.5-14B-Instruct2026.043.0146.942.412.34
SAMDModel=LLaMA-3.2-3B-Instruct2026.042.9972.792.182.17
SAMDModel=Qwen2.5-14B-Instruct2026.042.9345.862.352.19
PLDModel=Qwen2.5-7B-Instruct2026.042.3764.561.971.91
PLDModel=LLaMA-3.1-8B-Instruct2026.042.3560.722.152.12
PLDModel=LLaMA-3.2-3B-Instruct2026.042.1969.812.141.99
PLDModel=Qwen2.5-14B-Instruct2026.041.9835.31.811.82
VanillaModel=Qwen2.5-7B-Instruct2026.04132.7611
VanillaModel=Qwen2.5-14B-Instruct2026.04119.511
VanillaModel=LLaMA-3.1-8B-Instruct2026.04128.2111
VanillaModel=LLaMA-3.2-3B-Instruct2026.04132.611