Vision-Language Reasoning and Knowledge on MMMU
57.73AccuracyInternVL-3.5-8B
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| InternVL-3.5-8BModel Size=8B, Version=3.52026.05 | 57.73 | 1 | 26.45 | 36.28 | 34.04 | |
| InternVL-3.0-8BModel Size=8B, Version=3.02026.05 | 55.37 | 2 | 12.51 | 20.42 | 12.79 | |
| Qwen3.0-VL-8BModel Size=8B, Version=3.02026.05 | 54.9 | 3 | 22.9 | 32.32 | 40.13 | |
| Qwen2.5-VL-7BModel Size=7B, Version=2.52026.05 | 54.07 | 4 | 14.52 | 22.9 | 17.95 | |
| Qwen2.0-VL-7BModel Size=7B, Version=2.02026.05 | 52.54 | 5 | 11.45 | 18.81 | 12.63 | |
| Gemma3.0-12BModel Size=12B, Version=3.02026.05 | 51.36 | 6 | 10.63 | 17.61 | 12.16 | |
| InternVL-2.5-8BModel Size=8B, Version=2.52026.05 | 50.65 | 7 | 17.95 | 26.5 | 17.9 | |
| InternVL-2.0-8BModel Size=8B, Version=2.02026.05 | 45.45 | 8 | 15.7 | 23.34 | 16.34 | |
| LLaVA-1.5-13BModel Size=13B, Version=1.52026.05 | 38.25 | 9 | 8.38 | 13.75 | 8.51 | |
| Qwen-VL-7B-ChatModel Size=7B, Mode=Chat2026.05 | 37.9 | 10 | 3.54 | 6.48 | 2.81 | |
| LLaVA-1.6-13BModel Size=13B, Version=1.62026.05 | 37.07 | 11 | 16.06 | 22.42 | 20.77 |