Agentic performance on BFCL Multi-turn V4
58.5Base ScoreQwen3-235B
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Qwen3-235Bfull_name=Qwen3-235B-A22B-Instruct, setting=non-thinking, using_tools=true2026.01 | 58.5 | 47.5 | 35 | 54 | 49.7 | |
| SYNTHAGENT-14Bsetting=non-thinking, using_tools=true, training_strategy=synthetic training2026.01 | 57 | 46.5 | 31 | 46 | 46.3 | |
| ToolStar-14Bsetting=non-thinking, using_tools=true2026.01 | 56.5 | 35.5 | 29.5 | 39.5 | 35.7 | |
| SYNTHAGENT-8Bsetting=non-thinking, using_tools=true, training_strategy=synthetic training2026.01 | 54.5 | 45.5 | 33 | 37.5 | 42.9 | |
| ToolStar-8Bsetting=non-thinking, using_tools=true2026.01 | 52 | 38 | 22.5 | 30.5 | 31.7 | |
| Qwen3-32Bsetting=non-thinking, using_tools=true2026.01 | 50.5 | 43 | 30.5 | 33 | 36 | |
| Qwen3-14Bsetting=non-thinking, using_tools=true2026.01 | 40 | 34.5 | 26.5 | 26.5 | 30.6 |