Ministral-3-8B-Reasoning-2512: benchmark results
Mistral Ministral 3 8B Reasoning 2512 snapshot. Provider: Mistral. Released 2025-12-02. Access: Open.
Unified ELO 1523 ± 18, rank #694 of 1636 rated models, from 65 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| KorMedMCQA-V - Non-Stress Test | 57.2 | Accuracy (%) on the questions carrying the 54 fetal non-stre | 74 |
| KorMedMCQA-V - Ultrasound | 66.4 | Accuracy (%) on the questions carrying the 138 ultrasound im | 74 |
| LLM Stats (AIME 2024) | 86 | Score (%) | 70.5 |
| KorMedMCQA-V - CT | 62 | Accuracy (%) on the questions carrying the 336 CT images (Ko | 70 |
| KorMedMCQA-V - X-Ray | 64 | Accuracy (%) on the questions carrying the 586 X-ray images | 70 |
| KorMedMCQA-V - MRI | 71.6 | Accuracy (%) on the questions carrying the 40 MRI images (Ko | 68 |
| KorMedMCQA-V - Two Images | 62.9 | Accuracy (%) on the 426 questions with two images (KorMedMCQ | 68 |
| KorMedMCQA-V - Endoscopy | 59.8 | Accuracy (%) on the questions carrying the 122 endoscopy ima | 67 |
| KorMedMCQA-V - Three or More Images | 59 | Accuracy (%) on the 39 questions with three or more images ( | 67 |
| KorMedMCQA-V | 65.1 | Accuracy (%) on all 1,534 questions (KorMedMCQA-V: image-bas | 66 |
| KorMedMCQA-V - Peripheral Blood Smear | 68.2 | Accuracy (%) on the questions carrying the 49 peripheral blo | 66 |
| KorMedMCQA-V - Single Image | 66.2 | Accuracy (%) on the 1,069 questions with one image (KorMedMC | 64 |
Interactive version: theaggregate.ai/model?slug=ministral-3-8b-reasoning-2512 · How It Works · Data refreshed daily, snapshot 2026-10-11.