multimodal benchmark
MMMU-Pro Leaderboard
MMMU-Pro is a multi-discipline multimodal understanding benchmark that enhances MMMU by filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings.
Updated Sep 5, 2026
Models69
Model coverage69
MetricScore
EvidenceB
MMMU-Pro Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelGO | Score83.60% | Percentile100.00% | Participants69 | EvidenceC | Evaluated |
| Rank02 | ModelOP | Score83.20% | Percentile98.53% | Participants69 | EvidenceC | Evaluated |
| Rank03 | ModelOP | Score83.00% | Percentile97.06% | Participants69 | EvidenceC | Evaluated |
| Rank04 | ModelBY | Score82.70% | Percentile95.59% | Participants69 | EvidenceC | Evaluated |
| Rank05 | ModelAC | Score82.30% | Percentile94.12% | Participants69 | EvidenceC | Evaluated |
| Rank06 | ModelBY | Score82.20% | Percentile92.65% | Participants69 | EvidenceC | Evaluated |
| Rank07 | ModelMA | Score81.60% | Percentile91.18% | Participants69 | EvidenceC | Evaluated |
| Rank08 | ModelGO | Score81.20% | Percentile89.71% | Participants69 | EvidenceC | Evaluated |
| Rank09 | ModelOP | Score81.20% | Percentile88.24% | Participants69 | EvidenceC | Evaluated |
| Rank10 | ModelGO | Score81.00% | Percentile86.76% | Participants69 | EvidenceC | Evaluated |
| Rank11 | ModelOP | Score80.70% | Percentile85.29% | Participants69 | EvidenceC | Evaluated |
| Rank12 | ModelGO | Score80.50% | Percentile83.82% | Participants69 | EvidenceC | Evaluated |
| Rank13 | ModelME | Score80.40% | Percentile82.35% | Participants69 | EvidenceC | Evaluated |
| Rank14 | ModelMA | Score80.10% | Percentile80.88% | Participants69 | EvidenceC | Evaluated |
| Rank15 | ModelOP | Score79.50% | Percentile79.41% | Participants69 | EvidenceC | Evaluated |
| Rank16 | ModelAC | Score79.00% | Percentile77.94% | Participants69 | EvidenceC | Evaluated |
| Rank17 | ModelAC | Score78.80% | Percentile76.47% | Participants69 | EvidenceC | Evaluated |
| Rank18 | ModelMA | Score78.50% | Percentile75.00% | Participants69 | EvidenceC | Evaluated |
| Rank19 | ModelOP | Score78.40% | Percentile73.53% | Participants69 | EvidenceC | Evaluated |
| Rank20 | ModelOP | Score78.40% | Percentile72.06% | Participants69 | EvidenceC | Evaluated |
| Rank21 | ModelMI | Score78.10% | Percentile70.59% | Participants69 | EvidenceC | Evaluated |
| Rank22 | ModelXI | Score77.90% | Percentile69.12% | Participants69 | EvidenceC | Evaluated |
| Rank23 | ModelAN | Score77.30% | Percentile67.65% | Participants69 | EvidenceC | Evaluated |
| Rank24 | ModelGO | Score76.90% | Percentile66.18% | Participants69 | EvidenceC | Evaluated |
| Rank25 | ModelAC | Score76.90% | Percentile64.71% | Participants69 | EvidenceC | Evaluated |
| Rank26 | ModelGO | Score76.80% | Percentile63.24% | Participants69 | EvidenceC | Evaluated |
| Rank27 | ModelOP | Score76.60% | Percentile61.76% | Participants69 | EvidenceC | Evaluated |
| Rank28 | ModelOP | Score76.40% | Percentile60.29% | Participants69 | EvidenceC | Evaluated |
| Rank29 | ModelOP | Score76.00% | Percentile58.82% | Participants69 | EvidenceC | Evaluated |
| Rank30 | ModelAC | Score75.80% | Percentile57.35% | Participants69 | EvidenceC | Evaluated |
MMMU-Pro Highlights
The leading models and scores on this benchmark.
MMMU-Pro Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for MMMU-Pro
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is MMMU-Pro?
What MMMU-Pro measures and how its scores work.
MMMU-Pro is a multimodal benchmark for evaluating multi-discipline understanding and extends MMMU through a three-step process.
It measures multimodal understanding performance using Score, reported as a ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- MMMU-Pro
- 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 MMMU-Pro.
Which model scores highest on MMMU-Pro?
Gemini 3.5 Flash is currently ranked first with 83.60%.
What are the top three models on MMMU-Pro?
The current leaders are Gemini 3.5 Flash (83.60%), GPT-5.5 (83.20%), and GPT-5.6 Sol (83.00%).
Which MMMU-Pro model has the lowest official input price?
Qwen2.5-Omni-7B has the lowest matched official input price at $0.10 input / $0.40 output per 1M tokens.
Which models are fastest among MMMU-Pro results?
The fastest matched records are Llama 4 Maverick (307.30 tok/s via Groq), Gemini 3.5 Flash (210.94 tok/s via Google), and GPT-5.5 Instant (206.89 tok/s via OpenAI).
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.