can't wait to see ppl fine-tuning our mini models 🤠

Mistral Large 3 debuts as the #1 open source coding model on the lmarena.ai leaderboard. We'd love for you to try it! More on coding in a few days... 👀

Introducing the Devstral 2 coding model family. Two sizes, both open source. Also, meet Mistral Vibe, a native CLI, enabling end-to-end automation. 🧵

Congrats to the Mistral AI team on the launch of Devstral 2! 🚀 vLLM now delivers Day-0 support for the Devstral 2 Instruct models — optimized for agentic coding, deep codebase exploration, and multi-file editing at scale. Feel free to reach out 👇

Congrats to Mistral AI on the big launch 🚀 You can check out Mistral Vibe in Zed today – just download it and add your API key!

👀

Proud to share the first public model I worked on at Mistral AI. A decoder-only LLM optimized for creative writing, narrative generation, roleplay, and character-driven dialogue. Now live via API as labs-mistral-small-creative docs.mistral.ai/models/mistral…

Introducing Mistral OCR 3, a new frontier in document intelligence! 🧵👇

Very happy to announce the release of our latest Mistral OCR, which significantly outperforms existing solutions! A lot of effort was done to improve handwritten content, low quality scans, and complex tables & forms commonly found in enterprise documents.

🎄 We've shipped Skills for Vibe CLI: your dev expertise and rules, bundled and reusable across projects. Plus reasoning model support and native terminal themes. Try Vibe CLI now and happy shipping! → uv tool install mistral-vibe

new Mistral paper: Ministral 3!

new Ministral 3 tech report from Mistral AI, they train competitive small models on 1/3T tokens only. the secret? pruning + distillation distillation: > in pre-training they use Mistral Small 3.1 Instruct as a teacher for ALL variants (so not really cascade distillation, each

New Mistral AI paper: Ministral 3!

Incredible debugging story from Mistral AI: a memory leak in vLLM that standard tools couldn't detect. The hunt went from Python profilers → pmap → BPFtrace → GDB, finally tracing it to UCX's mmap hooks. Fix merged: github.com/vllm-project/v… Thanks for sharing! A must-read

Introducing Voxtral Transcribe 2, next-gen speech-to-text models by Mistral AI. State-of-the-art transcription, speaker diarization, sub-200ms real-time latency. Details in 🧵

Congrats to Mistral AI on releasing Voxtral Mini 4B Realtime! 🎉 Day-0 support in vLLM! A 4B streaming ASR model achieving <500ms latency while matching offline model accuracy, supporting 13 languages. vLLM's new Realtime API `/v1/realtime` provides audio streaming - optimized

Introducing Mistral AI's biggest hackathon ever! 📅 Feb 28 - Mar 1 🌍 Paris | London | NY | SF | Tokyo | Singapore | Sydney & online 48 hours. The best hackers. 🤝 Partners: Weights & Biases NVIDIA Amazon Web Services Iterate 🏆 $200K in prizes. Special awards from ElevenLabs Hugging Face

the six love languages are words of affirmation, quality time, gift giving, acts of service, physical touch, and.. ..machine learning. this year ML valentines have a new home on the internet <3