| Benchmark | Claude 3.7 Sonnet | Qwen3.7 Plus Preview |
|---|---|---|
| AA Agentic Index | 37 | 20.8 |
| AA Intelligence | 27.6 | 39.4 |
| AA-LCR | 62.3 | 69 |
| AA-Omniscience | -0.7 | 2.4 |
| arena_vision | 1195 | 1263 |
| Artificial Analysis Coding Index | 36.4 | 55.9 |
| critpt | 0.9 | 9.1 |
| gdpval | 27.4 | 22.2 |
| GPQA Diamond | 84.8 | 90.3 |
| HLE | 9.7 | 35.6 |
| IFBench | 48.3 | 79.1 |
| ifeval | 93.2 | 94.6 |
| mmmlu | 86.1 | 89 |
| MMMU-Pro | 60.1 | 80.5 |
| OmniScience Accuracy | 28.2 | 22.5 |
| OmniScience Non-Hallucination | 60 | 74.5 |
| scicode | 40.3 | 51.3 |
| SWE-bench Verified | 70.3 | 77.7 |
| Terminal-Bench Hard | 21.2 | 47 |
| τ²-Bench Telecom (AA run) | 54.7 | 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.