| Benchmark | Model | Rank | Results |
|---|---|---|---|
| computer-use-agents-on-ui-v-ui-vision | ZwZ2-27B | #2 | Accuracy: 66.66 |
| visual-question-answering-on-v-bench | ZwZ2-27B | #2 | Accuracy: 91.62 |
| visual-question-answering-vqa-on-erqa | ZwZ2-27B | #2 | Accuracy: 61.75 |
| spatial-reasoning-on-mmsi-bench | ZwZ2-27B | #4 | Accuracy: 37.1 |
| computer-use-agents-on-screenspot-pro | ZwZ2-27B | #6 | Accuracy: 70.59 |
| multimodal-reasoning-on-babyvision | ZwZ2-27B | #10 | Accuracy: 28.35 |
| vision-language-understanding-on-mmstar | ZwZ2-27B | #13 | Accuracy: 79.07 |