Ministral 3 3B — benchmark results
Mistral's 3B edge model from the Mistral 3 family, released under Apache 2.0 with image understanding and base, instruct and reasoning variants (December 2025). Provider: Mistral. Released 2025-12-02. Access: Open.
Unified ELO 1494 ± 10, rank #822 of 1776 rated models, from 336 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval Polish NLU - Polemo2 | 98.75 | Sentiment classification Score (%) | 100 |
| EuroEval Catalan NLU - MultiWikiQA CA | 78.35 | Reading comprehension Score (%) | 99.1 |
| EuroEval Estonian NLU - MultiWikiQA ET | 77.34 | Reading comprehension Score (%) | 98.8 |
| EuroEval Icelandic NLU - NQII | 62.37 | Reading comprehension Score (%) | 98.2 |
| EuroEval Catalan NLU - Guia CAT | 75.72 | Sentiment classification Score (%) | 98.1 |
| EuroEval German NLU - Germanquad | 68.77 | Reading comprehension Score (%) | 96.9 |
| EuroEval Finnish NLU - Tydiqa FI | 76.11 | Reading comprehension Score (%) | 96.2 |
| EuroEval Slovene NLU - Sentinews | 54.45 | Sentiment classification Score (%) | 96.1 |
| EuroEval Czech NLU - SQAD | 70.84 | Reading comprehension Score (%) | 95.6 |
| EuroEval German Summarization - Mlsum DE | 38.47 | Score (%) | 95.3 |
| EuroEval Catalan NLU - WikiANN CA | 74.81 | Named entity recognition Score (%) | 94.3 |
| EuroEval Portuguese NLU - SST-2 PT | 85.31 | Sentiment classification Score (%) | 94.1 |
Interactive version: theaggregate.ai/model?slug=ministral-3-3b · How the rankings work · Data refreshed daily, snapshot 2026-07-22.