Professional-level Knowledge Acquisition on VideoMMMU
83.6AccuracyGemini-2.5-Pro
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Gemini-2.5-ProSize=-, Evaluation Frames=642026.03 | 83.6 | — | |
| GPT-4oSize=-, Evaluation Frames=642026.03 | 61.2 | 57.8 | |
| Insight-V++Size=7B, Evaluation Frames=642026.03 | 57.6 | 54.2 | |
| Insight-VSize=7B, Evaluation Frames=642026.03 | 55.8 | 52.9 | |
| Gemini-1.5-ProSize=-, Evaluation Frames=642026.03 | 53.9 | — | |
| Video-R1Size=7B, Evaluation Frames=642026.03 | 52.4 | — | |
| Gemini-1.5-FlashSize=-, Evaluation Frames=642026.03 | 49.8 | — | |
| VideoChat-R1Size=7B, Evaluation Frames=642026.03 | 49.6 | — | |
| Qwen2.5-VLSize=7B, Evaluation Frames=642026.03 | 47.4 | 47.3 | |
| VideoLLaMA3Size=7B, Evaluation Frames=642026.03 | 47 | — | |
| VideoChat-FlashSize=7B, Evaluation Frames=642026.03 | 41.7 | — | |
| LLaVA-VideoSize=7B, Evaluation Frames=642026.03 | 36.1 | — | |
| LLaVA-OneVisionSize=7B, Evaluation Frames=642026.03 | 33.9 | — |