language benchmark
MMLU-ProX Leaderboard
MMLU-ProX is an extended version of MMLU-Pro with additional challenging multiple-choice questions across diverse academic and professional domains.
Updated Sep 5, 2026
Models32
Model coverage32
MetricScore
EvidenceB
MMLU-ProX Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score87.00% | Percentile100.00% | Participants32 | EvidenceC | Evaluated |
| Rank02 | ModelAC | Score85.40% | Percentile96.77% | Participants32 | EvidenceC | Evaluated |
| Rank03 | ModelAC | Score84.70% | Percentile93.55% | Participants32 | EvidenceC | Evaluated |
| Rank04 | ModelAC | Score84.70% | Percentile90.32% | Participants32 | EvidenceC | Evaluated |
| Rank05 | ModelNV | Score83.00% | Percentile87.10% | Participants32 | EvidenceC | Evaluated |
| Rank06 | ModelAC | Score82.20% | Percentile83.87% | Participants32 | EvidenceC | Evaluated |
| Rank07 | ModelAC | Score82.20% | Percentile80.65% | Participants32 | EvidenceC | Evaluated |
| Rank08 | ModelAC | Score81.00% | Percentile77.42% | Participants32 | EvidenceC | Evaluated |
| Rank09 | ModelAC | Score81.00% | Percentile74.19% | Participants32 | EvidenceC | Evaluated |
| Rank10 | ModelAC | Score80.60% | Percentile70.97% | Participants32 | EvidenceC | Evaluated |
| Rank11 | ModelAC | Score79.40% | Percentile67.74% | Participants32 | EvidenceC | Evaluated |
| Rank12 | ModelNV | Score79.36% | Percentile64.52% | Participants32 | EvidenceC | Evaluated |
| Rank13 | ModelAC | Score78.70% | Percentile61.29% | Participants32 | EvidenceC | Evaluated |
| Rank14 | ModelAC | Score77.80% | Percentile58.06% | Participants32 | EvidenceC | Evaluated |
| Rank15 | ModelAC | Score77.20% | Percentile54.84% | Participants32 | EvidenceC | Evaluated |
| Rank16 | ModelAC | Score76.70% | Percentile51.61% | Participants32 | EvidenceC | Evaluated |
| Rank17 | ModelAC | Score76.30% | Percentile48.39% | Participants32 | EvidenceC | Evaluated |
| Rank18 | ModelAC | Score76.10% | Percentile45.16% | Participants32 | EvidenceC | Evaluated |
| Rank19 | ModelAC | Score73.40% | Percentile41.94% | Participants32 | EvidenceC | Evaluated |
| Rank20 | ModelAC | Score71.50% | Percentile38.71% | Participants32 | EvidenceC | Evaluated |
| Rank21 | ModelAC | Score70.90% | Percentile35.48% | Participants32 | EvidenceC | Evaluated |
| Rank22 | ModelAC | Score70.70% | Percentile32.26% | Participants32 | EvidenceC | Evaluated |
| Rank23 | ModelAC | Score65.40% | Percentile29.03% | Participants32 | EvidenceC | Evaluated |
| Rank24 | ModelAC | Score65.00% | Percentile25.81% | Participants32 | EvidenceC | Evaluated |
| Rank25 | ModelNV | Score59.50% | Percentile22.58% | Participants32 | EvidenceC | Evaluated |
| Rank26 | ModelAC | Score59.40% | Percentile19.35% | Participants32 | EvidenceC | Evaluated |
| Rank27 | ModelAC | Score52.30% | Percentile16.13% | Participants32 | EvidenceC | Evaluated |
| Rank28 | ModelAC | Score34.60% | Percentile12.90% | Participants32 | EvidenceC | Evaluated |
| Rank29 | ModelGO | Score19.90% | Percentile9.68% | Participants32 | EvidenceC | Evaluated |
| Rank30 | ModelGO | Score19.90% | Percentile6.45% | Participants32 | EvidenceC | Evaluated |
MMLU-ProX Highlights
The leading models and scores on this benchmark.
MMLU-ProX Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for MMLU-ProX
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is MMLU-ProX?
What MMLU-ProX measures and how its scores work.
MMLU-ProX is a language benchmark built on the Massive Multitask Language Understanding benchmark framework.
MMLU-ProX measures language model performance using a Score reported as a ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- MMLU-ProX
- 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-ProX.
Which model scores highest on MMLU-ProX?
Qwen3.7 Max is currently ranked first with 87.00%.
What are the top three models on MMLU-ProX?
The current leaders are Qwen3.7 Max (87.00%), Qwen3.7-Plus (85.40%), and Qwen3.5-397B-A17B (84.70%).
Which MMLU-ProX model has the lowest official input price?
Nemotron 3 Super (120B A12B) has the lowest matched official input price at $0.20 input / $0.80 output per 1M tokens.
Which models are fastest among MMLU-ProX results?
The fastest matched records are Qwen3 VL 4B Thinking (8.77 tok/s via DeepInfra).
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.