# SimpleQA → MMLU-Pro

Calibration · https://benchgap.net/calibration/1425 · data retrieved 2026-10-29

MMLU-Pro is estimated from SimpleQA with a Michaelis–Menten + offset curve fitted on 7 models measured on both: y = 0.0000 + 0.8948·x / (0.03317 + x), R² = 0.75, cross-validated error 4.8 pp. It is used for 1 estimate.

| Estimated model | SimpleQA | MMLU-Pro | Source |
|---|---|---|---|
| [Gemini 2.0 Flash-Lite](https://benchgap.net/model/gemini-2-0-flash-lite.md) | 16.5% | 74.5% | estimated ± 4.8 pp, medium confidence |

## Fitted on

| Model | SimpleQA | MMLU-Pro |
|---|---|---|
| [DeepSeek-R1](https://benchgap.net/model/deepseek-r1.md) | 30.1% | 84.0% |
| [DeepSeek-R1-0528](https://benchgap.net/model/deepseek-r1-0528.md) | 27.8% | 85.0% |
| [DeepSeek V3](https://benchgap.net/model/deepseek-v3.md) | 24.9% | 75.9% |
| [Grok 3 [Beta]](https://benchgap.net/model/grok-3-beta.md) | 43.6% | 79.9% |
| [Kimi K2](https://benchgap.net/model/kimi-k2.md) | 31.0% | 81.1% |
| [Mistral Small 3.1](https://benchgap.net/model/mistral-small-3-1.md) | 10.4% | 66.8% |
| [Qwen3 235B 2507](https://benchgap.net/model/qwen3-235b-2507.md) | 54.3% | 83.0% |