Knowledge and Reasoning on MMLU Redux (Score)
93.3MMLU Redux ScoreQwen3.5
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3.5Parameter Count=35BA3B2026.05 | 93.3 | |
| Gemma4Parameter Count=26BA4B2026.05 | 91.91 | |
| Qwen3.5Parameter Count=9B2026.05 | 91.1 | |
| Qwen3VLParameter Count=30BA3B, Configuration Variant=Think2026.05 | 90.9 | |
| Qwen3-30B-A3BGroup=Larger, Generation Configuration=Standardized2026.03 | 90 | |
| Qwen3-14BGroup=Larger, Generation Configuration=Standardized2026.03 | 89.9 | |
| Gpt-oss-20b-highGroup=Larger, Generation Configuration=Standardized2026.03 | 89.6 | |
| SenseNova-U1Parameter Count=30BA3B, Configuration Variant=Think2026.05 | 89.44 | |
| Qwen3VLParameter Count=8B, Configuration Variant=Think2026.05 | 88.8 | |
| SenseNova-U1Parameter Count=8B, Configuration Variant=Think2026.05 | 87.61 | |
| EngGPT2-16B-A3BGroup=Comparable, Generation Configuration=Standardized2026.03 | 75.5 | |
| gemma-3-12b-itGroup=Larger, Generation Configuration=Standardized2026.03 | 75.3 | |
| gemma-2-9b-itGroup=Comparable, Generation Configuration=Standardized2026.03 | 74.7 | |
| Moonlight-16B-A3B-InstructGroup=Comparable, Generation Configuration=Standardized2026.03 | 70.1 | |
| Llama-3.1-8B-InstructGroup=Comparable, Generation Configuration=Standardized2026.03 | 68.8 |