MMLU-Pro: leaderboard, scores and methodology
A harder version of MMLU with ten answer options and more reasoning-heavy questions.
Top 10
Bars are scaled to the highest observed score. The leader sets the pace; the spread below it shows how much separation exists in this benchmark.
Leaderboard
| # | Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|---|
| 1 | GPT-5.2 Pro | OpenAI | 2000000 | $21.32 | 89.6 |
| 2 | GPT-5.1 Max | OpenAI | 2000000 | $20.50 | 87.9 |
| 3 | Claude Opus 4.5 Pro | Anthropic | 200000 | $20.98 | 89.1 |
| 4 | Claude Opus 4.5 Reasoning | Anthropic | 2000000 | $20.91 | 88.2 |
| 5 | Gemini 3 Pro Reasoning | Google DeepMind | 200000 | $19.69 | 87.5 |
| 6 | GPT-5.1 Reasoning | OpenAI | 400000 | $20.27 | 88.3 |
| 7 | GPT-5.2 | OpenAI | 2000000 | $19.71 | 88.0 |
| 8 | Claude Opus 4.5 Max | Anthropic | 1000000 | $20.90 | 87.8 |
| 9 | GPT-5.2 Vision | OpenAI | 1000000 | $19.03 | 87.4 |
| 10 | Gemini 3 Pro Max | Google DeepMind | 2000000 | $20.57 | 87.9 |
| 11 | Qwen3-Max Pro | Alibaba Cloud | 400000 | $18.87 | 86.2 |
| 12 | Grok 4 Reasoning | xAI | 2000000 | $19.67 | 87.8 |
| 13 | Gemini 3 Pro | Google DeepMind | 2000000 | $18.24 | 86.1 |
| 14 | Gemini 3 Pro Pro | Google DeepMind | 200000 | $19.80 | 87.3 |
| 15 | Qwen3-Max Reasoning | Alibaba Cloud | 2000000 | $18.12 | 85.9 |
| 16 | o4 Turbo | OpenAI | 2000000 | $19.34 | 86.6 |
| 17 | Claude Sonnet 4.5 | Anthropic | 400000 | $17.84 | 86.0 |
| 18 | Gemini 3 Flash Turbo | Google DeepMind | 200000 | $17.07 | 84.1 |
| 19 | Gemini 3 Flash Max | Google DeepMind | 1000000 | $17.43 | 85.0 |
| 20 | Grok 4 | xAI | 200000 | $17.90 | 86.5 |
| 21 | Grok 4 Max | xAI | 2000000 | $19.70 | 86.6 |
| 22 | o3 Reasoning | OpenAI | 200000 | $18.07 | 85.3 |
| 23 | o4 Pro | OpenAI | 1000000 | $19.37 | 87.3 |
| 24 | Claude Sonnet 4.5 Reasoning | Anthropic | 400000 | $19.48 | 87.7 |
| 25 | GPT-5.2 Max | OpenAI | 1000000 | $21.39 | 89.2 |
| 26 | GPT-4.1 Max | OpenAI | 400000 | $17.34 | 84.7 |
| 27 | GPT-4.1 Reasoning | OpenAI | 1000000 | $16.97 | 83.8 |
| 28 | GPT-5.2 Turbo | OpenAI | 1000000 | $20.73 | 87.4 |
| 29 | Gemini 3 Pro Vision | Google DeepMind | 200000 | $17.59 | 85.5 |
| 30 | Grok 3 Max | xAI | 2000000 | $17.66 | 85.9 |
| 31 | Claude Sonnet 4.5 Turbo | Anthropic | 1000000 | $18.37 | 85.9 |
| 32 | Gemini 3 Flash Reasoning | Google DeepMind | 128000 | $17.05 | 83.7 |
| 33 | Claude Opus 4.5 | Anthropic | 400000 | $19.42 | 87.7 |
| 34 | GPT-5.2 Reasoning | OpenAI | 1000000 | $20.98 | 87.5 |
| 35 | Gemini 3 Flash Pro | Google DeepMind | 400000 | $17.71 | 84.8 |
| 36 | Claude Opus 4.5 Vision | Anthropic | 200000 | $17.70 | 85.9 |
| 37 | DeepSeek-V3.2 | DeepSeek | 1000000 | $5.88 | 84.0 |
| 38 | GPT-4.1 Pro | OpenAI | 400000 | $17.56 | 85.2 |
| 39 | o3 | OpenAI | 400000 | $16.80 | 84.5 |
| 40 | Gemini 2.5 Pro Reasoning | Google DeepMind | 128000 | $16.81 | 83.8 |
| 41 | Qwen3-Max Turbo | Alibaba Cloud | 1000000 | $17.70 | 84.7 |
| 42 | Gemini 2.5 Pro Max | Google DeepMind | 200000 | $16.94 | 84.8 |
| 43 | Claude Haiku 4.5 Max | Anthropic | 200000 | $15.98 | 83.8 |
| 44 | Kimi K2 Instruct | Moonshot AI | 1000000 | $5.80 | 82.3 |
| 45 | Claude 3.7 Sonnet Reasoning | Anthropic | 128000 | $16.30 | 83.7 |
| 46 | Claude Sonnet 4.5 Pro | Anthropic | 1000000 | $18.73 | 86.7 |
| 47 | o4 Max | OpenAI | 2000000 | $19.51 | 86.2 |
| 48 | GPT-4.1 | OpenAI | 1000000 | $15.71 | 82.4 |
| 49 | GPT-5.1 Pro | OpenAI | 200000 | $19.77 | 87.8 |
| 50 | Claude Sonnet 4.5 Max | Anthropic | 1000000 | $19.90 | 87.2 |
Methodology
MMLU-Pro is included as a knowledge benchmark with a typical score range of 0 to 100. The values in this index are deterministic modelled estimates generated for comparison at scale. They are not official leaderboard results and should not be quoted as vendor benchmarks.
Each model receives a score derived from its capability tier, release generation and benchmark difficulty. Scores are normalised to a 0–100 scale so that models from different providers can be placed side by side. The average shown is the arithmetic mean across all 600 scored models.
Read the leaderboard as a relative map. A five-point gap is meaningful at the top and less meaningful in the middle of the distribution. Pair the score with price and context data before choosing a model.
How to use this leaderboard
- Use the top 10 to shortlist candidates, not to declare a winner.
- Check whether the leader is open weights or proprietary before assuming you can deploy it.
- Compare the score against input and output price on each model page.
- Look at the model's scores on related benchmarks before trusting a single number.
FAQ
What is MMLU-Pro?
A harder version of MMLU with ten answer options and more reasoning-heavy questions.
Who leads MMLU-Pro?
GPT-5.2 Pro from OpenAI leads this index with a score of 91.6, followed by GPT-5.1 Max at 91.3.
What is a good MMLU-Pro score?
The top quartile starts around 80.9. The index average is 73.5, so a score above 80.9 is a strong result relative to this dataset.
How are MMLU-Pro scores calculated?
The values are modelled estimates on a 0–100 scale, generated from capability tier, release generation and benchmark difficulty. They are not official leaderboard results.
Which models are included?
All 600 models in the index are scored on MMLU-Pro, across 40 providers and every model category.