Tool-use agent performance on BFCL V3
79.5Avg@4SkillX
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SkillXModel=GLM-4.6, Experience extraction model alignment=Aligned with inference model2026.04 | 79.5 | 86 | |
| ExpeLModel=GLM-4.6, Experience extraction model alignment=Aligned with inference model2026.04 | 78.83 | 85.33 | |
| AWMModel=GLM-4.6, Experience extraction model alignment=Aligned with inference model2026.04 | 77.17 | 84 | |
| No MemoryModel=GLM-4.6, Experience extraction model alignment=Aligned with inference model2026.04 | 76.67 | 83.33 | |
| A-MemModel=GLM-4.6, Experience extraction model alignment=Aligned with inference model2026.04 | 76.5 | 83 | |
| SkillXModel=Kimi-K2-Instruct-0905, Experience extraction model alignment=GLM-4.6 used for experience extraction2026.04 | 66.83 | 81.33 | |
| ExpeLModel=Kimi-K2-Instruct-0905, Experience extraction model alignment=Aligned with inference model2026.04 | 66.33 | 79.33 | |
| ExpeLModel=Kimi-K2-Instruct-0905, Experience extraction model alignment=GLM-4.6 used for experience extraction2026.04 | 66 | 79.67 | |
| AWMModel=Kimi-K2-Instruct-0905, Experience extraction model alignment=Aligned with inference model2026.04 | 65.33 | 79 | |
| No MemoryModel=Kimi-K2-Instruct-0905, Experience extraction model alignment=Aligned with inference model2026.04 | 65.17 | 78 | |
| A-MemModel=Kimi-K2-Instruct-0905, Experience extraction model alignment=Aligned with inference model2026.04 | 65.17 | 76.67 | |
| AWMModel=Kimi-K2-Instruct-0905, Experience extraction model alignment=GLM-4.6 used for experience extraction2026.04 | 64.67 | 79.17 | |
| SkillXModel=Qwen3-32B, Experience extraction model alignment=GLM-4.6 used for experience extraction2026.04 | 63.67 | 82 | |
| ExpeLModel=Qwen3-32B, Experience extraction model alignment=GLM-4.6 used for experience extraction2026.04 | 59.33 | 78.83 | |
| ExpeLModel=Qwen3-32B, Experience extraction model alignment=Aligned with inference model2026.04 | 57.33 | 77.67 | |
| AWMModel=Qwen3-32B, Experience extraction model alignment=GLM-4.6 used for experience extraction2026.04 | 56.67 | 76.33 | |
| AWMModel=Qwen3-32B, Experience extraction model alignment=Aligned with inference model2026.04 | 55.67 | 76 | |
| No MemoryModel=Qwen3-32B, Experience extraction model alignment=Aligned with inference model2026.04 | 53.67 | 73.33 | |
| A-MemModel=Qwen3-32B, Experience extraction model alignment=Aligned with inference model2026.04 | 53.67 | 73 |