| Benchmark | Model | Rank | Results |
|---|---|---|---|
| multimodal-reasoning-on-mmmu | DeepAlign (InternVL3-8B) | – | Accuracy: 63.7 |
| optical-character-recognition-ocr-on-ocrbench | DeepAlign (InternVL3-8B) | – | Score: 890 |
| vision-language-understanding-on-mmstar | DeepAlign (InternVL3-8B) | – | Accuracy: 69.8 |
| visual-question-answering-vqa-on-realworldqa | DeepAlign (InternVL3-8B) | – | Accuracy: 71.6 |