Function Calling on BFCL Simple (Live)
80.1AccuracyQwen3-8B
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3-8BModel Type=Autoregressive LLM, Format=Python2026.07 | 80.1 | |
| Llama-3.1-8BModel Type=Autoregressive LLM, Format=Python2026.07 | 70.8 | |
| Dream + Log-depth samplingModel Type=Diffusion LLM (dLLM), Decoding Method=Log-depth sampling, Format=Python2026.07 | 69.7 | |
| LLaDA + Log-depth samplingModel Type=Diffusion LLM (dLLM), Decoding Method=Log-depth sampling, Format=Python2026.07 | 64 | |
| LLaDAModel Type=Diffusion LLM (dLLM), Decoding Method=Unconstrained, Format=Python2026.07 | 50.3 | |
| DreamModel Type=Diffusion LLM (dLLM), Decoding Method=Unconstrained, Format=Python2026.07 | 22.4 |