| Benchmark | Model | Rank | Results |
|---|---|---|---|
| mathematical-reasoning-on-mathverse | Qwen3-VL-Instruct-8B + GCPO | #1 | Accuracy: 84.1 |
| multimodal-reasoning-on-mathvision | Qwen3-VL-Instruct-8B + GCPO | #26 | Accuracy: 56.9 |
| text-to-image-generation-on-geneval | Janus-Pro-R1-7B + GCPO | #33 | Overall: 0.89 |
| multimodal-reasoning-on-mmmu-pro | Qwen3-VL-Instruct-8B + GCPO | #36 | Accuracy: 55.3 |