| Benchmark | o4-mini | Qwen3.7 Plus Preview |
|---|---|---|
| AA Agentic Index | 36.1 | 20.8 |
| AA Intelligence | 26.1 | 39.4 |
| AA-LCR | 60 | 69 |
| AA-Omniscience | -35.7 | 2.4 |
| arena_vision | 1201 | 1263 |
| Artificial Analysis Coding Index | 25.6 | 55.9 |
| critpt | 0.6 | 9.1 |
| gdpval | 25.4 | 22.2 |
| GPQA Diamond | 81.4 | 90.3 |
| HLE | 16.5 | 35.6 |
| IFBench | 68.7 | 79.1 |
| mmlu_prox | 69.3 | 85.4 |
| MMMU-Pro | 69.2 | 80.5 |
| OmniScience Accuracy | 24.8 | 22.5 |
| OmniScience Non-Hallucination | 19.5 | 74.5 |
| scicode | 46.5 | 51.3 |
| SWE-bench Verified | 68.1 | 77.7 |
| Terminal-Bench Hard | 15.2 | 47 |
| τ²-Bench Telecom (AA run) | 55.6 | 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.