| Benchmark | Claude 4 Opus | Qwen3.7 Plus Preview |
|---|---|---|
| AA Intelligence | 31.7 | 39.4 |
| AA-LCR | 40 | 69 |
| arena_vision | 1207 | 1263 |
| Artificial Analysis Coding Index | 34 | 55.9 |
| GPQA Diamond | 79.6 | 90.3 |
| HLE | 12.3 | 35.6 |
| IFBench | 53.7 | 79.1 |
| ifeval | 87.4 | 94.6 |
| LiveCodeBench v6 | 47.4 | 89.6 |
| mmlu_redux | 94.2 | 94.5 |
| MMLU-Pro | 86.6 | 88.5 |
| mmmlu | 88.8 | 89 |
| scicode | 40.9 | 51.3 |
| supergpqa | 56.5 | 71.4 |
| SWE-bench Verified | 79.4 | 77.7 |
| Terminal-Bench Hard | 31.1 | 47 |
| τ²-Bench Telecom (AA run) | 73.4 | 93 |
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.