Benchmark
MMVU
MMVU (Multimodal Multi-disciplinary Video Understanding) is a benchmark for evaluating multimodal models on video understanding tasks across multiple…
- Modality
- multimodal
- Categories
- multimodal, reasoning, video, vision
- Openness
- unknown
- Source
- llm_stats
- Reported scores
- 4
Reported scores
llm_stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Kimi K2.5 | Moonshot AI | 0.804 | 2026-01-27 | evidence |
| Qwen3.5-122B-A10B | Qwen | 0.747 | 2026-02-24 | evidence |
| Qwen3.5-27B | Qwen | 0.733 | 2026-02-24 | evidence |
| Qwen3.5-35B-A3B | Qwen | 0.723 | 2026-02-24 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.