MMEB Video: leaderboard
MMEB video split evaluating multimodal embedding models on video classification, QA, retrieval, and multi-video retrieval tasks.
Metric: Video Overall Score (%). Source: huggingface.co. 101 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | octen-vl-embedding-large | 75.95 | |
| 2 | DME-Large | 74.36 | |
| 3 | WeMM-Embedding-9B | 74.31 | |
| 4 | Ovis-VL-Embedding-9B | 72.9 | |
| 5 | WeMM-Embedding-4B | 72.11 | |
| 6 | Wianmu-5B-v0 | 71.56 | |
| 7 | WeMM-Embedding-2B | 70.82 | |
| 8 | Wianmu-3B-v0 | 70.79 | |
| 9 | DME-Medium | 70.79 | |
| 10 | Wanderboat-Embedding-2B | 70.71 |
Interactive version: theaggregate.ai/benchmark?slug=mmeb-video · How It Works · Data refreshed daily, snapshot 2026-10-20.