Video-MME
Video understanding across 6 domains, clips from 11 s to 1 h, with subtitles and audio
multimodalactive
Definition
- Version
- not in the notes
- Category
- multimodal
- Maintainer
- Video-MME team (academic, China)
- Status
- active
- Home page
- https://video-mme.github.io/home_page.html
Sources
- Video-MME video-mme.github.io https://video-mme.github.io/home_page.html
Scores
| Model | Score | Who ran it | Setting | View or population | Date | Source |
|---|---|---|---|---|---|---|
 video-SALMONN 2+ 72B | 81.6 % | maintainer-run | With subtitles | rank 1, Video-MME leaderboard | 28 Sep 2025 | video-mme.github.io |
Where this comes from
Scores from different settings, harnesses or dates are not directly comparable; the row says which. Sorted by score, highest first, as a reading aid, not a ranking.
To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Video-MME", https://aurion.semai.id/benchmarks/video-mme/, records of October 2026.