| Benchmark | Claude 3.5 Sonnet | Qwen3.7 Max |
|---|---|---|
| AA Intelligence | 10 | 46.7 |
| Artificial Analysis Coding Index | 30.2 | 66 |
| GPQA Diamond | 67.2 | 92.4 |
| HLE | 3.9 | 41.4 |
| ifeval | 90.1 | 94.3 |
| LiveCodeBench v6 | 37.2 | 91.6 |
| longbench_v2 | 41 | 65.3 |
| mmlu_redux | 88.9 | 95 |
| MMLU-Pro | 78 | 89.6 |
| scicode | 36.6 | 53.5 |
| simplebench | 41.4 | 70.4 |
| supergpqa | 48.2 | 73.6 |
| SWE-bench Verified | 50.8 | 80.4 |
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.