Mistral Medium (2604): benchmark results
Provider: Mistral. Released 2023-12-11. Access: API.
Aggregate Rating 1631 ± 34, rank #345 of 1640 rated models, from 14 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| JudgeSense - Coherence | 0.89 | Judge Sensitivity Score: share of items on which the judge r | 100 |
| JudgeSense - Relevance | 0.95 | Judge Sensitivity Score: share of items on which the judge r | 62.5 |
| JudgeSense - Preference | 0.93 | Judge Sensitivity Score: share of items on which the judge r | 54.5 |
| Guesswork 2026-07 | 1.04 | MAE (z-score units) | 44.4 |
| Guesswork 2026-09 | 0.89 | MAE (z-score units) | 44.4 |
| Guesswork 2026-08 | 1.21 | MAE (z-score units) | 43.8 |
| Epoch AI - Dtbench | 75.47 | Score | 39 |
| JudgeSense - Factuality | 0.89 | Judge Sensitivity Score: share of items on which the judge r | 31.2 |
| Conceptual Reasoning Index - Argument Evaluation (LMCA) | 26.06 | Chance-Corrected Score (0-100) | 30.9 |
| Guesswork 2026-10 | 1.23 | MAE (z-score units) | 18.8 |
| VAmoS Pro Bench | 22.7 | Task Completion (%) | 12.5 |
| Epoch AI - Critpt | 0 | Score | 9.6 |
Interactive version: theaggregate.ai/model?slug=mistral-medium-2604 · How It Works · Data refreshed daily, snapshot 2026-10-16.