| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| MVBench | LLaVA-Video-7B | Accuracy100 | 425 | 3d ago | |
| VideoMME | Gemini-1.5-Pro | Score (Long)67.4 | 248 | 11d ago | |
| VideoMME | Overall Score100 | 222 | 11d ago | ||
| MLVU | Score78.19 | 221 | 11d ago | ||
| EgoSchema | FLoC | EgoSchema Score69.4 | 158 | 4d ago | |
| MVBench (test) | FSR | Accuracy100.3 | 151 | 15d ago | |
| Video-MME | HiMu | Overall Score78.18 | 96 | 24d ago | |
| Video-MME | Overall Score78.78 | 92 | 18d ago | ||
| LongVideoBench | VisionZip† + DyTok (7B) | LongVideoBench Score59.2 | 92 | 4d ago | |
| LVB | FLoC | Accuracy66.49 | 89 | 1mo ago | |
| Video-MME without subtitles | Gemini-1.5-Pro | Overall Score75 | 89 | 1mo ago | |
| MLVU | Accuracy87.34 | 80 | 22d ago | ||
| LongVideoBench, MLVU, and VideoMME Aggregate | LLaVA-OneVision-7B | Average Score56.4 | 75 | 1mo ago | |
| LVBench | Average Score73.5 | 67 | 4d ago | ||
| VideoMME | Accuracy (No Subtitles)65.1 | 60 | 10d ago | ||
| Aggregate MVBench, LongVideo Bench, MLVU, VideoMME | LLaVA-Video-7B | Average Score100 | 59 | 5d ago | |
| Video-MME v1.0 (test) | TS-LLaVA | Score (Short)72.4 | 56 | 1mo ago | |
| TempCompass MCQ (test) | Accuracy82.8 | 55 | 18d ago | ||
| EgoSchema (test) | Qwen2-VL-72B | Accuracy77.9 | 55 | 29d ago | |
| Video-MME (test) | Accuracy88.6 | 51 | 15d ago | ||
| VideoEvalPro (test) | Gemini 2.5 Pro | Accuracy0.784 | 50 | 18d ago | |
| MotionBench (val) | Accuracy65.4 | 50 | 18d ago | ||
| MLVU 3-120min (test) | LLaVA-OneVision-7B | Accuracy47.7 | 49 | 1mo ago | |
| MLVU 3-120min (dev) | LLaVA-OneVision-7B | Accuracy63 | 49 | 1mo ago | |
| LongVideoBench 1-60min | CaCoVID | Accuracy56.8 | 49 | 1mo ago |