Mistral Large: benchmark results
Mistral AI's first-generation Large, an API-only flagship (February 2024). Provider: Mistral. Released 2024-02-26. Access: API.
Aggregate Rating 1521 ± 8, rank #710 of 1656 rated models, from 376 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenCompass Code - HumanEval - Chinese (CompassBench 2407) | 83.5 | Score (%) | 100 |
| OpenCompass Code - HumanEval+ - English (CompassBench 2407) | 81.7 | Score (%) | 100 |
| OpenCompass Knowledge - Humanities - Chinese (CompassBench 2407) | 79.2 | Score (%) | 100 |
| OpenCompass Agent - CIBench - Chinese (CompassBench 2405) | 75.2 | Score (%) | 97.4 |
| OpenCompass Agent - CIBench - English (CompassBench 2405) | 76.8 | Score (%) | 97.4 |
| OpenCompass LLM - Agent (CompassBench 2405) | 80.8 | Score (%) | 97.4 |
| OpenCompass LLM - Agent - Chinese (CompassBench 2405) | 80.1 | Score (%) | 97.4 |
| OpenCompass LLM - Agent - English (CompassBench 2405) | 81.5 | Score (%) | 97.4 |
| OpenCompass Knowledge - Science - English (CompassBench 2407) | 91.2 | Score (%) | 97.1 |
| OpenCompass Language - NLP - English (CompassBench 2407) | 55.5 | Score (%) | 97.1 |
| OpenCompass Math - College - Chinese (CompassBench 2407) | 69 | Score (%) | 97.1 |
| OpenCompass LLM - Knowledge - English (CompassBench 2405) | 82.9 | Score (%) | 94.7 |
Interactive version: theaggregate.ai/model?slug=mistral-large · How It Works · Data refreshed daily, snapshot 2026-10-20.