| Benchmark | Model | Rank | Results |
|---|---|---|---|
| chart-question-answering-on-chartqa | TowerAlign | – | 1:1 Accuracy: 59.0 |
| multimodal-reasoning-on-mmmu | TowerAlign | – | Accuracy: 40.3 |
| vision-language-understanding-on-mmstar | TowerAlign | – | Accuracy: 46.7 |
| visual-question-answering-vqa-on-realworldqa | TowerAlign | – | Accuracy: 58.9 |