language benchmark
MMLU-Redux Leaderboard
MMLU-Redux is an improved version of the MMLU benchmark with manually re-annotated questions to identify and correct errors in the original dataset.
Updated Sep 7, 2026
Models48
Model coverage48
MetricScore
EvidenceB
MMLU-Redux Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score95.00% | Percentile100.00% | Participants48 | EvidenceC | Evaluated |
| Rank02 | ModelAC | Score94.90% | Percentile97.87% | Participants48 | EvidenceC | Evaluated |
| Rank03 | ModelAC | Score94.50% | Percentile95.74% | Participants48 | EvidenceC | Evaluated |
| Rank04 | ModelAC | Score94.50% | Percentile93.62% | Participants48 | EvidenceC | Evaluated |
| Rank05 | ModelMA | Score94.40% | Percentile91.49% | Participants48 | EvidenceC | Evaluated |
| Rank06 | ModelAC | Score94.00% | Percentile89.36% | Participants48 | EvidenceC | Evaluated |
| Rank07 | ModelAC | Score93.80% | Percentile87.23% | Participants48 | EvidenceC | Evaluated |
| Rank08 | ModelAC | Score93.70% | Percentile85.11% | Participants48 | EvidenceC | Evaluated |
| Rank09 | ModelAC | Score93.50% | Percentile82.98% | Participants48 | EvidenceC | Evaluated |
| Rank10 | ModelDE | Score93.40% | Percentile80.85% | Participants48 | EvidenceC | Evaluated |
| Rank11 | ModelAC | Score93.30% | Percentile78.72% | Participants48 | EvidenceC | Evaluated |
| Rank12 | ModelAC | Score93.30% | Percentile76.60% | Participants48 | EvidenceC | Evaluated |
| Rank13 | ModelAC | Score93.20% | Percentile74.47% | Participants48 | EvidenceC | Evaluated |
| Rank14 | ModelAC | Score93.10% | Percentile72.34% | Participants48 | EvidenceC | Evaluated |
| Rank15 | ModelXI | Score92.80% | Percentile70.21% | Participants48 | EvidenceC | Evaluated |
| Rank16 | ModelMA | Score92.70% | Percentile68.09% | Participants48 | EvidenceC | Evaluated |
| Rank17 | ModelMA | Score92.70% | Percentile65.96% | Participants48 | EvidenceC | Evaluated |
| Rank18 | ModelAC | Score92.50% | Percentile63.83% | Participants48 | EvidenceC | Evaluated |
| Rank19 | ModelAC | Score92.20% | Percentile61.70% | Participants48 | EvidenceC | Evaluated |
| Rank20 | ModelAC | Score91.90% | Percentile59.57% | Participants48 | EvidenceC | Evaluated |
| Rank21 | ModelDE | Score91.80% | Percentile57.45% | Participants48 | EvidenceC | Evaluated |
| Rank22 | ModelAC | Score91.10% | Percentile55.32% | Participants48 | EvidenceC | Evaluated |
| Rank23 | ModelAC | Score90.90% | Percentile53.19% | Participants48 | EvidenceC | Evaluated |
| Rank24 | ModelAC | Score90.90% | Percentile51.06% | Participants48 | EvidenceC | Evaluated |
| Rank25 | ModelAC | Score89.80% | Percentile48.94% | Participants48 | EvidenceC | Evaluated |
| Rank26 | ModelME | Score89.30% | Percentile46.81% | Participants48 | EvidenceC | Evaluated |
| Rank27 | ModelDE | Score89.10% | Percentile44.68% | Participants48 | EvidenceC | Evaluated |
| Rank28 | ModelAC | Score88.80% | Percentile42.55% | Participants48 | EvidenceC | Evaluated |
| Rank29 | ModelAC | Score88.80% | Percentile40.43% | Participants48 | EvidenceC | Evaluated |
| Rank30 | ModelAC | Score88.40% | Percentile38.30% | Participants48 | EvidenceC | Evaluated |
MMLU-Redux Highlights
The leading models and scores on this benchmark.
MMLU-Redux Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for MMLU-Redux
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is MMLU-Redux?
What MMLU-Redux measures and how its scores work.
MMLU-Redux is a language benchmark that addresses dataset quality issues found in the original MMLU.
It measures language model performance using the Score metric, expressed as a ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- MMLU-Redux
- 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-Redux.
Which model scores highest on MMLU-Redux?
Qwen3.7 Max is currently ranked first with 95.00%.
What are the top three models on MMLU-Redux?
The current leaders are Qwen3.7 Max (95.00%), Qwen3.5-397B-A17B (94.90%), and Qwen3.6 Plus (94.50%).
Which MMLU-Redux model has the lowest official input price?
Ministral 3 (14B Base 2512) has the lowest matched official input price at $0.10 input / $0.10 output per 1M tokens.
Which models are fastest among MMLU-Redux results?
The fastest matched records are DeepSeek-V3 (100.00 tok/s via DeepSeek), DeepSeek-R1-0528 (45.04 tok/s via DeepInfra), and Qwen2.5-Coder 32B Instruct (44.00 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.