Fine-grained Visual Reasoning on MME-RW
70.2ScoreQwen3.5
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3.5Model Category=General Large Vision-Language Models, Model Size=9B, Independent Testing Condition=true2026.06 | 70.2 | |
| V-Zero-4BModel Category=Visually Grounded Reasoning Models, Model Size=4B2026.06 | 69.8 | |
| ZwZModel Category=Visually Grounded Reasoning Models, Model Size=8B, Independent Testing Condition=true2026.06 | 69.6 | |
| Qwen3.5Model Category=General Large Vision-Language Models, Model Size=4B, Independent Testing Condition=true2026.06 | 69.2 | |
| ZwZModel Category=Visually Grounded Reasoning Models, Model Size=4B, Independent Testing Condition=true2026.06 | 68.5 | |
| DeepEyesV2Model Category=Visually Grounded Reasoning Models, Model Size=7B2026.06 | 64.9 | |
| ThymeModel Category=Visually Grounded Reasoning Models, Model Size=7B2026.06 | 64.8 | |
| Pixel-ReasonerModel Category=Visually Grounded Reasoning Models, Model Size=7B2026.06 | 64.4 | |
| DeepEyesModel Category=Visually Grounded Reasoning Models, Model Size=7B2026.06 | 64.1 | |
| Qwen3-VLModel Category=General Large Vision-Language Models, Model Size=4B, Independent Testing Condition=true2026.06 | 63.5 |