multimodal benchmark
VideoMME w sub.
The first-ever comprehensive evaluation benchmark of Multi-modal LLMs in Video analysis. Features 900 videos (254 hours) with 2,700 question-answer pairs covering 6 primary visual domains and 30 subfields. Evaluates temporal understanding across short (11 seconds) to long (1 hour) videos with multi-modal inputs including video frames, subtitles, and audio.
Updated Aug 7, 2026
Published models10
Registry coverage10
MetricScore
EvidenceB
VideoMME w sub. leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | AC | 90.4% | 100.0% | 10 | C | |
| 2 | AC | 87.7% | 88.9% | 10 | C | |
| 3 | AC | 87.3% | 77.8% | 10 | C | |
| 4 | AC | 87.0% | 66.7% | 10 | C | |
| 5 | OP | 86.7% | 55.6% | 10 | C | |
| 6 | AC | 86.6% | 44.4% | 10 | C | |
| 7 | AC | 86.6% | 33.3% | 10 | C | |
| 8 | AC | 77.9% | 22.2% | 10 | C | |
| 9 | AC | 72.4% | 11.1% | 10 | C | |
| 10 | AC | 71.6% | 0.0% | 10 | C |
Score distribution
Top published rows on the benchmark's original scale.
VideoMME w sub. highlights
The top published results on this benchmark's own scale.
What is VideoMME w sub.?
Definition and scoring fields from the benchmark registry.
The first-ever comprehensive evaluation benchmark of Multi-modal LLMs in Video analysis. Features 900 videos (254 hours) with 2,700 question-answer pairs covering 6 primary visual domains and 30 subfields. Evaluates temporal understanding across short (11 seconds) to long (1 hour) videos with multi-modal inputs including video frames, subtitles, and audio.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- VideoMME w sub.
- Modality
- multimodal
- Primary category
- multimodal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- videomme-w-sub.|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about VideoMME w sub..
Which model scores highest on VideoMME w sub.?
Qwen3.8 Max is currently ranked first with 90.4%.
What does VideoMME w sub. measure?
The first-ever comprehensive evaluation benchmark of Multi-modal LLMs in Video analysis. Features 900 videos (254 hours) with 2,700 question-answer pairs covering 6 primary visual domains and 30 subfields. Evaluates temporal understanding across short (11 seconds) to long (1 hour) videos with multi-modal inputs including video frames, subtitles, and audio.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
10 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.