Ministral-3-14B-Instruct-2512: benchmark results
Mistral Ministral 3 14B Instruct 2512 snapshot. Provider: Mistral. Released 2025-12-02. Access: Open.
Aggregate Rating 1550 ± 10, rank #615 of 1656 rated models, from 440 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| SPUR - Heterogeneous Integration | 70 | Accuracy (%) on SPUR the 130 heterogeneous integration multi | 100 |
| SPUR - Trend Analysis | 57.79 | Accuracy (%) on SPUR the 1,357 trend analysis multiple-choic | 100 |
| EuroEval Lithuanian NLU - Atsiliepimai | 50.64 | Matthews correlation coefficient (%) | 99.5 |
| EuroEval French NLU - Allocine | 96.93 | Matthews correlation coefficient (%) | 98.3 |
| EuroEval Faroese NLU - FoQA | 78.86 | Character F1 (%) | 97.7 |
| EuroEval French NLU - ScaLA FR | 70.67 | Matthews correlation coefficient (%) | 97.6 |
| EuroEval Italian NLU - SQuAD IT | 75.14 | Character F1 (%) | 96.7 |
| EuroEval Portuguese NLU - SST-2 PT | 87.08 | Matthews correlation coefficient (%) | 96.1 |
| EuroEval Ukrainian NLU - ScaLA UK | 43.85 | Matthews correlation coefficient (%) | 96 |
| EuroEval French NLU | 73.63 | NLU Average Score (%) | 95.2 |
| JP-TL-Bench - Japanese to English - Easy | 9.9 | LT Score (0-10; Easy items only; anchored pairwise Bradley-T | 95.2 |
| EuroEval Estonian NLU - MultiWikiQA ET | 73.2 | Character F1 (%) | 94 |
Interactive version: theaggregate.ai/model?slug=ministral-3-14b-instruct-2512 · How It Works · Data refreshed daily, snapshot 2026-10-20.