List your company

Tool-use Reasoning on BFCL V4

18Average PerformanceLlama-3.1-8B-Instruct + RODS (ours)
6.6129.568512.52515.4815Jun 17, 2026

Evaluation results

MethodLinks
Llama-3.1-8B-Instruct + RODS (ours)Training=RODS2026.061815.520.5
Llama-3.1-8B-Instruct + EnvTuningTraining=EnvTuning2026.0616.531518.06
ToolACE-2-8B2026.0615.758.922.6
xLAM-2-8b-fc-r2026.069.17513.33
BitAgent-8B2026.067.414.510.32
Llama-3.1-8B-Instruct2026.067.052.511.6