OCR and Chart Understanding on OCRBench
881Total ScoreQwen3-VL-4B
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3-VL-4BModel Scale=4B2026.02 | 881 | |
| Qwen3-VLLanguage Backbone=Qwen3 1.7B LLM, Evaluation Framework=VLMEvalKit [69]2026.03 | 869 | |
| FineViT-VLLanguage Backbone=Qwen3 1.7B LLM, Evaluation Framework=VLMEvalKit [69]2026.03 | 841 | |
| Intern3.5-VLLanguage Backbone=Qwen3 1.7B LLM, Evaluation Framework=VLMEvalKit [69]2026.03 | 835 | |
| InternVL-3.5-4BModel Scale=4B2026.02 | 822 | |
| DenseMLLM-4BModel Scale=4B2026.02 | 813 | |
| Aquila-VLLanguage Backbone=Qwen2.5 1.5B LLM, Evaluation Framework=VLMEvalKit [69]2026.03 | 772 | |
| Qwen2.5-VL-7B +PCBackbone Model=Qwen2.5-VL-7B, Training Paradigm (+PC)=Yes2026.03 | 91.1 | |
| Qwen2.5-VL-7BBackbone Model=Qwen2.5-VL-7B, Training Paradigm (+PC)=No2026.03 | 88.8 | |
| Qwen2.5-VLOpenness=semi-open, Model Size=7B2025.10 | 86.4 | |
| Keye-VLOpenness=semi-open, Model Size=8B2025.10 | 85.1 | |
| MiMo-VL-7B +PCBackbone Model=MiMo-VL-7B, Training Paradigm (+PC)=Yes2026.03 | 84.8 | |
| InternVL3.5Openness=semi-open, Model Size=8B2025.10 | 84 | |
| Bee-8BOpenness=fully open, Model Size=8B, Training Strategy=SFT2025.10 | 83.1 | |
| MiMo-VL-7BBackbone Model=MiMo-VL-7B, Training Paradigm (+PC)=No2026.03 | 82.9 | |
| Bee-8BOpenness=fully open, Model Size=8B, Training Strategy=RL2025.10 | 82.5 | |
| Qwen2.5-VL-3B +PCBackbone Model=Qwen2.5-VL-3B, Training Paradigm (+PC)=Yes2026.03 | 81.5 | |
| Qwen2.5-VL-3BBackbone Model=Qwen2.5-VL-3B, Training Paradigm (+PC)=No2026.03 | 79.7 | |
| MolmoOpenness=fully open, Model Size=7B, Training Strategy=Distilled (-D)2025.10 | 65.6 | |
| LLaVA OneVisionOpenness=fully open, Model Size=7B2025.10 | 62.2 |