This model is over 2 years old. A newer version of it has since shipped: Mistral OCR 4 (2026-06-23). The details below are still accurate for this model — it just is not what we would recommend today.

Mistral Large 2

Mistral AI🇫🇷 France
deprecated

Mistral's top model. Strong coding, multilingual, and instruction following.

Context window131K tokens
Input / 1M tokens$2
Output / 1M tokens$6

Version History

mistral-large-2409minor

Mistral Large 2 September update with improved code generation and expanded function calling support for complex tool schemas.

mistral-large-2407major

Mistral Large 2 with 128K context, major coding gains, and multilingual improvements.

Benchmark Scores

Full leaderboard →
67.8%
Hallucination Rate
76.4%
HumanEval
81.2%
MMLU
72.0 tokens_per_sec
Speed (tok/s)