| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| VideoMME v1 (test) | Accuracy77.2 | 27 | 2mo ago | ||
| MLVU v1 (test) | CoRDS | Accuracy71.5 | 26 | 2mo ago | |
| EgoSchema v1 (test) | Accuracy72.2 | 22 | 2mo ago | ||
| VideoMME w/o sub. | FluxMem | Accuracy (Overall)65.3 | 16 | 4mo ago | |
| LongVideo-Bench | GPT-4o | Accuracy66.7 | 13 | 1mo ago | |
| VideoMME | GPT-4o | Accuracy71.9 | 12 | 1mo ago | |
| Video-MME | Accuracy (with subtitle)67.6 | 4 | 4mo ago |