| Benchmark | Gemini 1.5 Pro | Qwen3.6 35B A3B |
|---|---|---|
| AA Intelligence | 10 | 32.1 |
| AI2D | 80.9 | 92.7 |
| Artificial Analysis Coding Index | 23.6 | 41.9 |
| GPQA Diamond | 59.1 | 86 |
| HLE | 4.9 | 22.2 |
| MathVista | 70.6 | 86.4 |
| MMLU-Pro | 75.8 | 85.6 |
| MMMU | 68.4 | 81.7 |
| MMMU-Pro | 55 | 75.3 |
| ocrbench_v2 | 51.6 | 65.5 |
| scicode | 29.5 | 35.8 |
| SWE-bench Verified | 34.2 | 73.4 |
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.