Mistral Small 3.2: benchmark results
Mistral Small 3.2, a 24B refinement of Small 3.1 (June 2025). Provider: Mistral. Released 2025-06-20. Access: Open.
Aggregate Rating 1553 ± 6, rank #605 of 1656 rated models, from 906 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Cedar - Multimodal - Spanish | 48.15 | Accuracy (%) | 100 |
| EuroEval Ukrainian NLU | 60.27 | NLU Average Score (%) | 100 |
| Vibe-Bench - BERT Retrieval - Response BERT-F1 | 10.7 | Score (0-100) | 100 |
| Vibe-Bench - BM25 Retrieval - Response BERT-F1 | 10.81 | Score (0-100) | 100 |
| Vibe-Bench - BM25 Retrieval - Strategy Macro-F1 | 18.91 | Score (0-100) | 100 |
| Vibe-Bench - BM25 Retrieval - Strategy Weighted-F1 | 22.55 | Score (0-100) | 100 |
| Vibe-Bench - Full History - Preference Accuracy | 69.53 | Score (0-100) | 100 |
| Vibe-Bench - Full History - Preference F1 | 69.29 | Score (0-100) | 100 |
| Vibe-Bench - Full History - Preference MCC | 0.39 | Matthews Correlation (-1 to 1) | 100 |
| Vibe-Bench - Full History - Preference Precision | 69.84 | Score (0-100) | 100 |
| Vibe-Bench - Full History - Response BERT-F1 | 9.44 | Score (0-100) | 100 |
| Vibe-Bench - Full History - Strategy Balanced Accuracy | 19.87 | Score (0-100) | 100 |
Interactive version: theaggregate.ai/model?slug=mistral-small-3-2 · How It Works · Data refreshed daily, snapshot 2026-10-20.