Function Calling on Berkeley Function Call Leaderboard (BFCL) online inference setting
621.13Input TokensQwen3-8B
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Qwen3-8BModel Category=Public Models (General)2026.02 | 621.13 | 54.68 | 1.984 | |
| Qwen3-4BModel Category=Public Models (General)2026.02 | 621.13 | 58.24 | 1.341 | |
| Qwen3-1.7BModel Category=Public Models (General)2026.02 | 621.13 | 55.26 | 1.014 | |
| Qwen3-0.6BModel Category=Public Models (General)2026.02 | 621.13 | 53.58 | 1.006 | |
| Qwen2.5-14B-InstructModel Category=Public Models (General)2026.02 | 632.82 | 57.95 | 2.985 | |
| Qwen2.5-7B-InstructModel Category=Public Models (General)2026.02 | 632.82 | 58.99 | 1.815 | |
| Qwen2.5-3B-InstructModel Category=Public Models (General)2026.02 | 632.82 | 62.7 | 1.282 | |
| Qwen2.5-1.5B-InstructModel Category=Public Models (General)2026.02 | 632.82 | 56.63 | 0.872 | |
| Qwen2.5-0.5B-InstructModel Category=Public Models (General)2026.02 | 632.82 | 60.54 | 0.927 | |
| HyFunc (Func. Call. Gen.)LML=ToolACE-8B, LMS=Qwen3-0.6B2026.02 | 702.25 | 36.67 | 0.724 | |
| HyFunc (Function Retriever)LML=ToolACE-8B, LMS=Qwen3-0.6B2026.02 | 708.51 | 1 | 0.104 | |
| xLAM-7B-rModel Category=Public Models (Specifical)2026.02 | 810.57 | 59.65 | 1.171 | |
| Hammer2.1-7BModel Category=Public Models (Specifical)2026.02 | 822 | 57.74 | 1.792 | |
| Hammer2.1-3BModel Category=Public Models (Specifical)2026.02 | 822 | 52.98 | 1.042 | |
| Hammer2.1-1.5BModel Category=Public Models (Specifical)2026.02 | 822 | 56.48 | 0.949 | |
| Hammer2.1-0.5BModel Category=Public Models (Specifical)2026.02 | 822 | 52.54 | 0.904 | |
| xLAM-1B-rModel Category=Public Models (Specifical)2026.02 | 864.03 | 65.47 | 0.911 | |
| ToolACE-8BModel Category=Public Models (Specifical)2026.02 | 943.05 | 55.34 | 1.876 | |
| HyFunc (Total)LML=ToolACE-8B, LMS=Qwen3-0.6B2026.02 | 1,410.76 | 37.67 | 0.828 |