Mistral 7B Instruct (v0.1): benchmark results
Provider: Mistral. Released 2023-09-27. Access: Open.
Unified ELO 1361 ± 1, rank #1243 of 1392 rated models, from 130 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| NPHardEval - GCP | 20.91 | Accuracy (%) | 100 |
| Enkrypt AI - Bias Risk | 51.94 | Risk Score | 91.7 |
| NPHardEval - GCP D | 58 | Accuracy (%) | 90.9 |
| NPHardEval - TSP D | 62.73 | Accuracy (%) | 81.8 |
| Open Medical LLM - PubMedQA | 75.8 | Accuracy (%) | 78.4 |
| MERA - RWSD | 61.15 | Accuracy (%) | 71.3 |
| SpeechMap Compliance | 70.4 | % Requests Completed | 63.1 |
| FastEval | 42.66 | Total Score | 62.5 |
| PIQA | 82.2 | Accuracy (%) | 60.8 |
| MERA - ruHumanEval | 17.8 | pass@1 (%) | 60.6 |
| EuroEval Portuguese NLU - MultiWikiQA PT | 70.47 | Reading comprehension Score (%) | 60.1 |
| MERA - ruCodeEval | 16.95 | pass@1 (%) | 59.3 |
Interactive version: theaggregate.ai/model?slug=mistral-7b-instruct-v0-1 · How It Works · Data refreshed daily, snapshot 2026-09-05.