| Benchmark | Model | Rank | Results |
|---|---|---|---|
| mathematical-reasoning-on-mathverse | CoM (Qwen2.5-VL-7B + Qwen2.5-Math-7B) | – | Accuracy: 50.25 |
| multimodal-reasoning-on-mathvision | CoM (Qwen2.5-VL-72B + Qwen2.5-Math-72B) | – | Accuracy: 40.95 |
| multimodal-reasoning-on-mathvista | CoM (Qwen2.5-VL-72B + Qwen2.5-Math-72B) | – | Accuracy: 75.9 |
| vision-language-understanding-on-mmstar | CoM (Qwen2.5-VL-7B + DeepSeek-R1-Distill-Qwen-7B) | – | Accuracy: 65.0 |