language benchmark
MMLU-Pro Leaderboard
MMLU-Pro is a multi-task language understanding benchmark that extends MMLU with 10 multiple-choice options, over 12,000 curated questions across 14 domains, and reasoning-intensive tasks.
Updated Sep 5, 2026
Models100
Model coverage138
MetricScore
EvidenceB
MMLU-Pro Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelSA | Score90.33% | Percentile100.00% | Participants138 | EvidenceC | Evaluated |
| Rank02 | ModelAC | Score89.60% | Percentile99.27% | Participants138 | EvidenceC | Evaluated |
| Rank03 | ModelAC | Score88.50% | Percentile98.54% | Participants138 | EvidenceC | Evaluated |
| Rank04 | ModelAC | Score88.50% | Percentile97.81% | Participants138 | EvidenceC | Evaluated |
| Rank05 | ModelMI | Score88.00% | Percentile97.08% | Participants138 | EvidenceC | Evaluated |
| Rank06 | ModelAC | Score87.80% | Percentile96.35% | Participants138 | EvidenceC | Evaluated |
| Rank07 | ModelDE | Score87.50% | Percentile95.62% | Participants138 | EvidenceC | Evaluated |
| Rank08 | ModelMA | Score87.10% | Percentile94.89% | Participants138 | EvidenceC | Evaluated |
| Rank09 | ModelBA | Score87.00% | Percentile94.16% | Participants138 | EvidenceC | Evaluated |
| Rank10 | ModelNV | Score86.80% | Percentile93.43% | Participants138 | EvidenceC | Evaluated |
| Rank11 | ModelAC | Score86.70% | Percentile92.70% | Participants138 | EvidenceC | Evaluated |
| Rank12 | ModelDE | Score86.40% | Percentile91.97% | Participants138 | EvidenceC | Evaluated |
| Rank13 | ModelUP | Score86.30% | Percentile91.24% | Participants138 | EvidenceC | Evaluated |
| Rank14 | ModelDE | Score86.20% | Percentile90.51% | Participants138 | EvidenceC | Evaluated |
| Rank15 | ModelAC | Score86.20% | Percentile89.78% | Participants138 | EvidenceC | Evaluated |
| Rank16 | ModelAC | Score86.10% | Percentile89.05% | Participants138 | EvidenceC | Evaluated |
| Rank17 | ModelAC | Score85.30% | Percentile88.32% | Participants138 | EvidenceC | Evaluated |
| Rank18 | ModelGO | Score85.20% | Percentile87.59% | Participants138 | EvidenceC | Evaluated |
| Rank19 | ModelAC | Score85.20% | Percentile86.86% | Participants138 | EvidenceC | Evaluated |
| Rank20 | ModelDE | Score85.00% | Percentile86.13% | Participants138 | EvidenceC | Evaluated |
| Rank21 | ModelDE | Score85.00% | Percentile85.40% | Participants138 | EvidenceC | Evaluated |
| Rank22 | ModelDE | Score85.00% | Percentile84.67% | Participants138 | EvidenceC | Evaluated |
| Rank23 | ModelDE | Score85.00% | Percentile83.94% | Participants138 | EvidenceC | Evaluated |
| Rank24 | ModelMI | Score85.00% | Percentile83.21% | Participants138 | EvidenceC | Evaluated |
| Rank25 | ModelXI | Score84.90% | Percentile82.48% | Participants138 | EvidenceC | Evaluated |
| Rank26 | ModelZA | Score84.60% | Percentile81.75% | Participants138 | EvidenceC | Evaluated |
| Rank27 | ModelMA | Score84.60% | Percentile81.02% | Participants138 | EvidenceC | Evaluated |
| Rank28 | ModelAC | Score84.40% | Percentile80.29% | Participants138 | EvidenceC | Evaluated |
| Rank29 | ModelZA | Score84.30% | Percentile79.56% | Participants138 | EvidenceC | Evaluated |
| Rank30 | ModelLA | Score83.80% | Percentile78.83% | Participants138 | EvidenceC | Evaluated |
MMLU-Pro Highlights
The leading models and scores on this benchmark.
MMLU-Pro Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for MMLU-Pro
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is MMLU-Pro?
What MMLU-Pro measures and how its scores work.
MMLU-Pro is a language benchmark extending MMLU by expanding multiple-choice options from 4 to 10 and eliminating trivial questions.
It reports Score as a ratio for performance on questions across 14 domains.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- MMLU-Pro
- Modality
- text
- Primary category
- language
- 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 MMLU-Pro.
Which model scores highest on MMLU-Pro?
Sakana Namazu is currently ranked first with 90.33%.
What are the top three models on MMLU-Pro?
The current leaders are Sakana Namazu (90.33%), Qwen3.7 Max (89.60%), and Qwen3.6 Plus (88.50%).
Which MMLU-Pro model has the lowest official input price?
Phi 4 Reasoning Plus has the lowest matched official input price at $0.13 input / $0.50 output per 1M tokens.
Which models are fastest among MMLU-Pro results?
The fastest matched records are Llama 3.3 70B Instruct (2,220.00 tok/s via Cerebras), Llama 4 Scout (776.10 tok/s via Groq), and Llama 4 Maverick (307.30 tok/s via Groq).
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.