| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| LongVideoBench | LLaVA-Video-7B | Score58.9 | 269 | 19d ago | |
| LongVideoBench (val) | LVAgent | Accuracy80 | 225 | 26d ago | |
| LVBench | Accuracy77 | 218 | 21h ago | ||
| MLVU | Qwen3-VL-235B-A22B | Accuracy83.8 | 205 | 21h ago | |
| LongVideo Bench | Qwen3-VL-8B-Instruct | Score62.8 | 99 | 21d ago | |
| LongVideoBench | Penguin-VL | Accuracy67 | 97 | 6d ago | |
| Video-MME Long | AVP | Accuracy81.9 | 92 | 15d ago | |
| VideoMME | Accuracy81.3 | 89 | 2d ago | ||
| MLVU (dev) | Qwen2.5-VL+AdaRETAKE | Score78.1 | 63 | 1mo ago | |
| MLVU (test) | Symphony | Average Score81 | 60 | 1mo ago | |
| VideoNIAH | Video-Zero | Accuracy46.81 | 54 | 19d ago | |
| LSDBench | InternVL3.5-4B | Accuracy60.35 | 54 | 19d ago | |
| Video-MME Overall | Accuracy87 | 53 | 1mo ago | ||
| Video-MME (full) | Qwen2.5-VL-128 | Overall Performance66.4 | 51 | 12d ago | |
| Video-MME | Gemini-2.5-Pro | Overall Score84.3 | 48 | 14d ago | |
| Video-MME long 1.0 | Gemini 1.5 Pro | Accuracy (No Subs)67.4 | 45 | 3mo ago | |
| LVBench (test) | Symphony | LVBench Score71.8 | 43 | 2mo ago | |
| MLVU 3-120 min | Accuracy82.1 | 36 | 1mo ago | ||
| LVOmniVideo | SEATS | Score38.5 | 32 | 14d ago | |
| Video MME w/o sub (long) | LensWalk | Accuracy71.4 | 30 | 1mo ago | |
| Long Video Bench | Score (15s)66.67 | 28 | 2mo ago | ||
| MLVU v1.0 (test) | FLoC | MLVU Score67.77 | 28 | 1mo ago | |
| VideoMME Long split, 30-60 min | Accuracy65.3 | 27 | 1mo ago | ||
| LongVideoBench | FastVID | LongVideoBench Score57.1 | 24 | 5d ago | |
| Video-MME (w/o sub.) Overall 1010s | Gemini 1.5 Pro | Accuracy75 | 22 | 1mo ago |