multimodal benchmark
MMMU-Pro (with tools)
MMMU-Pro variant evaluated with tool access enabled.
Updated Aug 7, 2026
Published models4
Registry coverage4
MetricScore
EvidenceB
MMMU-Pro (with tools) leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | OP | 84.6% | 100.0% | 4 | C | |
| 2 | MA | 83.4% | 66.7% | 4 | C | |
| 3 | OP | 82.0% | 33.3% | 4 | C | |
| 4 | OP | 79.5% | 0.0% | 4 | C |
Score distribution
Top published rows on the benchmark's original scale.
MMMU-Pro (with tools) highlights
The top published results on this benchmark's own scale.
What is MMMU-Pro (with tools)?
Definition and scoring fields from the benchmark registry.
MMMU-Pro variant evaluated with tool access enabled.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- MMMU-Pro (with tools)
- Modality
- multimodal
- Primary category
- multimodal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- mmmu-pro-with-tools|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about MMMU-Pro (with tools).
Which model scores highest on MMMU-Pro (with tools)?
GPT-5.6 Sol is currently ranked first with 84.6%.
What does MMMU-Pro (with tools) measure?
MMMU-Pro variant evaluated with tool access enabled.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
4 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.