Benchmark
MLVU Benchmark
What is MLVU?
A comprehensive benchmark for multi-task long video understanding that evaluates multimodal large language models on videos ranging from 3 minutes to 2 hours…
- Released
- 2024-06-06
- Modality
- multimodal
- Evaluates
- long context, multimodal, video, vision
- Openness
- unknown
- Importer
- LLM Stats
- Reported scores
- 10
Source provenance
- Original evidence https://api.zeroeval.com/leaderboard/benchmarks/mlvu?top_n=500
MLVU results and reported scores
LLM Stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Qwen2.5 VL 7B Instruct | Qwen | 0.702 | 2025-01-26 | evidence |
| Qwen3 VL 235B A22B Instruct | Qwen | 0.843 | 2025-09-22 | evidence |
| Qwen3 VL 235B A22B Thinking | Qwen | 0.838 | 2025-09-22 | evidence |
| Qwen3.5-122B-A10B | Qwen | 0.873 | 2026-02-24 | evidence |
| Qwen3.5-27B | Qwen | 0.859 | 2026-02-24 | evidence |
| Qwen3.5-35B-A3B | Qwen | 0.856 | 2026-02-24 | evidence |
| Qwen3.6 Plus | Qwen | 0.867 | 2026-04-02 | evidence |
| Qwen3.6-27B | Qwen | 0.866 | 2026-04-21 | evidence |
| Qwen3.6-35B-A3B | Qwen | 0.862 | 2026-04-16 | evidence |
| Qwen3.7-Plus | Qwen | 0.874 | 2026-05-31 | 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.