Multi-discipline Understanding on MMBench en (dev)
85.9AccuracyLLaVA-OneVision-72B
Evaluation results
| Method | Links | |
|---|---|---|
| LLaVA-OneVision-72BModel Scale=72B2024.08 | 85.9 | |
| Qwen3-VL-8BMode=base2026.05 | 84.4 | |
| S^3-FTBackbone=Qwen3-VL-8B, Mode=Single-view2026.05 | 81.9 | |
| S^3-FTBackbone=Qwen2.5-VL-7B, Mode=Single-view2026.05 | 81.7 | |
| S^3-FTBackbone=Qwen3-VL-8B, Mode=Cross-viewpoint2026.05 | 81.3 | |
| LLaVA-OneVision-7BModel Scale=7B2024.08 | 80.8 | |
| Qwen2.5-VL-7BMode=base2026.05 | 80.5 | |
| S^3-FTBackbone=Qwen2.5-VL-7B, Mode=Cross-viewpoint2026.05 | 80 | |
| GPT-4VVersion=V-Preview2024.08 | 75 | |
| LLaVA-OneVision-0.5BModel Scale=0.5B2024.08 | 52.1 |