| Benchmark | Claude 3.5 Sonnet | Kimi K2.6 |
|---|---|---|
| AA Intelligence | 10 | 45.1 |
| arena_vision | 1146 | 1264 |
| Artificial Analysis Coding Index | 30.2 | 61.8 |
| Chinese SimpleQA (C-SimpleQA) | 56.8 | 75.9 |
| frontiermath_tier_4 | 0 | 14.6 |
| GPQA Diamond | 67.2 | 91.1 |
| HLE | 3.9 | 37.5 |
| LiveCodeBench | 32.8 | 89.6 |
| LiveCodeBench v6 | 37.2 | 89.6 |
| MMLU-Pro | 78 | 87.1 |
| MMMU-Pro | 54.7 | 80.1 |
| scicode | 36.6 | 53.5 |
| SWE-bench Verified | 50.8 | 80.2 |
Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing.