Across 12 shared benchmarks, Gemini 1.5 Pro scores higher on 8 and Qwen2.5 Omni 7B on 4. The widest gap is GPQA Diamond, where Gemini 1.5 Pro scores 59.1 against 30.8.
| Benchmark | Gemini 1.5 Pro | Qwen2.5 Omni 7B |
|---|---|---|
| AI2D | 80.9 | 83.2 |
| chartqa | 88.7 | 85.3 |
| docvqa | 91.5 | 95.2 |
| FLEURS | 93.3 | 95.9 |
| GPQA Diamond | 59.1 | 30.8 |
| GSM8K | 95.2 | 88.7 |
| humaneval | 86.6 | 78.7 |
| MathVista | 70.6 | 67.9 |
| MMLU-Pro | 75.8 | 47 |
| MMMU | 68.4 | 59.2 |
| MMMU-Pro | 55 | 36.6 |
| ocrbench_v2 | 51.6 | 57.8 |
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.