Announcing Magistral — our Mistral AI first reasoning model — excelling in domain-specific, transparent, and multilingual reasoning.

- We’re releasing the model in two variants: Magistral Small — a 24B parameter open-source version and Magistral Medium — a more powerful, enterprise version. - Magistral Medium scored 73.6% on AIME2024, and 90% with majority voting @64. Magistral Small scored 70.7% and 83.3% respectively. - Reason natively — Magistral’s chain-of-thought works across global languages and alphabets. - Suited for a wide range of enterprise use cases — from structured calculations and programmatic logic to decision trees and rule-based systems.

@64 Multilingual dexterity: The model excels in maintaining high-fidelity reasoning across numerous languages. Magistral is especially well-suited to reason in languages including English, French, Spanish, German, Italian, Arabic, Russian, and Simplified Chinese.

@64 10x faster reasoning with Le Chat: With Flash Answers in Le Chat, Magistral Medium achieves up to 10x faster token throughput than most competitors. This enables real-time reasoning and user feedback, at scale.

@64 The release is supported by our latest paper covering comprehensive evaluations of Magistral, our training infrastructure, reinforcement learning algorithm, and novel observations for training reasoning models.

Magistral Small is an open-weight model, and is available for self-deployment under the Apache 2.0 license. You can use it with the following frameworks: - vllm (recommended): The community has prepared quantized versions of the model that can be used with the following frameworks (alphabetically sorted): - llama.cpp: - @lmstudio (llama.cpp, MLX): - @ollama (llama.cpp): - @UnslothAI (llama.cpp): Fine-tuning is possible with: - @axolotl_ai: - @UnslothAI: Also you can use Magistral with: - @kaggle:

@64 You can try out a preview version of Magistral Medium in Le Chat ( or via API on La Plateforme.

@64 Check out our blog post for details:

KSM achieves 5 ISO certifications! A major step for secure, AI-driven healthcare innovation. 🚀 #DataPrivacy #HealthTech

@MistralAI No base model on HF?

@MistralAI amazing work sophia!

@MistralAI Excellent news, Sophia.