Language Understanding on MMLU (Accuracy and Token Counts)
92.63AccuracyVanilla MAS + MOC
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Vanilla MAS + MOCAgent Backbone=DeepSeek-V3.2-685B-Chat2026.06 | 92.63 | 2,405,945 | 486,600 | |
| Vanilla MASAgent Backbone=DeepSeek-V3.2-685B-Chat2026.06 | 92.28 | 1,958,918 | 490,362 | |
| Single LLMAgent Backbone=DeepSeek-V3.2-685B-Chat2026.06 | 87.72 | 52,904 | 1,140 |