multimodal benchmark
MMMU (validation)
Validation set of the Massive Multi-discipline Multimodal Understanding and Reasoning benchmark. Features college-level multimodal questions across 6 core disciplines (Art & Design, Business, Science, Health & Medicine, Humanities & Social Science, Tech & Engineering) spanning 30 subjects and 183 subfields with diverse image types including charts, diagrams, maps, and tables.
Updated Aug 11, 2026
Models4
Model coverage4
MetricScore
EvidenceB
MMMU (validation) Ranking
Higher score ranks better on this benchmark.
| 01 | AN | 80.7% | 100.0% | 4 | C | |
| 02 | AN | 77.1% | 66.7% | 4 | C | |
| 03 | AN | 76.5% | 33.3% | 4 | C | |
| 04 | AN | 73.2% | 0.0% | 4 | C |
MMMU (validation) Score Distribution
A closer view of the leading scores on this benchmark.
MMMU (validation) Highlights
The leading models and scores on this benchmark.
What is MMMU (validation)?
What MMMU (validation) measures and how its scores work.
Validation set of the Massive Multi-discipline Multimodal Understanding and Reasoning benchmark. Features college-level multimodal questions across 6 core disciplines (Art & Design, Business, Science, Health & Medicine, Humanities & Social Science, Tech & Engineering) spanning 30 subjects and 183 subfields with diverse image types including charts, diagrams, maps, and tables.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- MMMU (validation)
- Modality
- multimodal
- Primary category
- multimodal
- Score direction
- higher
- LLMBoard eligible
- Yes
- Evaluation key
- mmmu-(validation)|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about MMMU (validation).
Which model scores highest on MMMU (validation)?
Claude Opus 4.5 is currently ranked first with 80.7%.
What does MMMU (validation) measure?
Validation set of the Massive Multi-discipline Multimodal Understanding and Reasoning benchmark. Features college-level multimodal questions across 6 core disciplines (Art & Design, Business, Science, Health & Medicine, Humanities & Social Science, Tech & Engineering) spanning 30 subjects and 183 subfields with diverse image types including charts, diagrams, maps, and tables.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
4 model results are currently shown.
Does this benchmark affect the overall score?
Yes. This benchmark can contribute to the current LLMBoard capability score.