multimodal benchmark
MMBench-Video Leaderboard
MMBench-Video is a long-form multi-shot benchmark for holistic video understanding that includes approximately 600 YouTube videos across 16 major categories, with videos ranging from 30 seconds to 6 minutes and roughly 2,000 original question-answer pairs.
Updated Sep 5, 2026
Models3
Model coverage3
MetricScore
EvidenceB
MMBench-Video Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score2.02% | Percentile100.00% | Participants3 | EvidenceC | Evaluated |
| Rank02 | ModelAC | Score1.93% | Percentile50.00% | Participants3 | EvidenceC | Evaluated |
| Rank03 | ModelAC | Score1.79% | Percentile0.00% | Participants3 | EvidenceC | Evaluated |
MMBench-Video Highlights
The leading models and scores on this benchmark.
MMBench-Video Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for MMBench-Video
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is MMBench-Video?
What MMBench-Video measures and how its scores work.
MMBench-Video is a multimodal benchmark comprising approximately 600 web videos from YouTube and roughly 2,000 original question-answer pairs.
It measures holistic video understanding across 26 fine-grained capabilities using the Score metric, reported as a ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- MMBench-Video
- Modality
- multimodal
- Primary category
- multimodal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about MMBench-Video.
Which model scores highest on MMBench-Video?
Qwen2.5 VL 72B Instruct is currently ranked first with 2.02%.
What are the top three models on MMBench-Video?
The current leaders are Qwen2.5 VL 72B Instruct (2.02%), Qwen2.5 VL 32B Instruct (1.93%), and Qwen2.5 VL 7B Instruct (1.79%).
Which MMBench-Video model has the lowest official input price?
Qwen2.5 VL 7B Instruct has the lowest matched official input price at $0.35 input / $1.1 output per 1M tokens.
Which models are fastest among MMBench-Video results?
No matched runtime record is currently available.
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.