Tool-Augmented Planning on BFCL v3
84.1Live Success RateMagicAgent-32B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MagicAgent-32BModel Category=MagicAgent Series2026.02 | 84.1 | 89.7 | |
| MagicAgent-30B-A3BModel Category=MagicAgent Series2026.02 | 83.3 | 89.6 | |
| Qwen3-MAXModel Category=Ultra-Scale2026.02 | 82.5 | 89.1 | |
| Kimi-K2-InstructModel Category=Ultra-Scale2026.02 | 82.1 | 82.5 | |
| GLM-4.7Model Category=Ultra-Scale2026.02 | 80.8 | 86.6 | |
| Qwen3-235B-A22B-InstructModel Category=Ultra-Scale2026.02 | 80.1 | 87.2 | |
| Qwen3-32B-nothinkModel Category=Large-Scale2026.02 | 80.1 | 86.3 | |
| GLM-4.7-FlashModel Category=Large-Scale2026.02 | 77.3 | 82.9 | |
| GPT-5.2Model Category=Ultra-Scale2026.02 | 76.8 | 82.2 | |
| Qwen3-30B-A3B-InstructModel Category=Large-Scale2026.02 | 75.8 | 85.3 | |
| DeepSeek-V3.1-nothinkModel Category=Ultra-Scale2026.02 | 71.1 | 67.6 | |
| ERNIE-4.5-21B-A3B-PTModel Category=Large-Scale2026.02 | 63.2 | 70.6 | |
| Llama3.3-70B-InstructModel Category=Large-Scale2026.02 | 62.1 | 79.9 | |
| Olmo-3.1-32B-InstructModel Category=Large-Scale2026.02 | 44.4 | 42.4 |