Baptiste Rozière (@b_roziere) 's Twitter Profile
Baptiste Rozière

@b_roziere

Research scientist at Mistral AI. Working on ML for programming languages. Lead author of CodeLlama, now working on the Codestral models!

ID: 781777461268709376

calendar_today30-09-2016 08:47:26

120 Tweet

1,1K Followers

253 Following

Mistral AI (@mistralai) 's Twitter Profile Photo

magnet:?xt=urn:btih:11f2d1ca613ccf5a5c60104db9f3babdfa2e6003&dn=Mistral-Small-3-Instruct&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=http%3A%2F%https://t.co/ua2yzvEYLu%3A1337%2Fannounce

Mistral AI (@mistralai) 's Twitter Profile Photo

Introducing Small 3, our most efficient and versatile model yet! Pre-trained and instructed version, Apache 2.0, 24B, 81% MMLU, 150 tok/s. No synthetic data so great base for anything reasoning - happy building! mistral.ai/news/mistral-s…

Mistral AI (@mistralai) 's Twitter Profile Photo

Le Chat is fast (1,100 tok/s for flash queries on an updated Mistral Large). Download it at mistral.ai/app/android or mistral.ai/app/ios

Sophia Yang, Ph.D. (@sophiamyang) 's Twitter Profile Photo

Announcing Mistral AI OCR -  the world’s best document understanding API. 🔍 State-of-the-art understanding of complex documents 🌍 Natively multilingual and multimodal ⚡ Fastest in its category 📄 Doc-as-prompt, structured output 🔒 Available for on-prem deployment

Announcing <a href="/MistralAI/">Mistral AI</a> OCR -  the world’s best document understanding API.

🔍 State-of-the-art understanding of complex documents
🌍 Natively multilingual and multimodal
⚡ Fastest in its category
📄 Doc-as-prompt, structured output
🔒 Available for on-prem deployment
Pierre Chambon (@pierrechambon6) 's Twitter Profile Photo

Does your LLM truly comprehend the complexity of the code it generates? 🥰   Introducing our new non-saturated (for at least the coming week? 😉) benchmark:   ✨BigO(Bench)✨ - Can LLMs Generate Code with Controlled Time and Space Complexity?   Check out the details below !👇

Does your LLM truly comprehend the complexity of the code it generates? 🥰
 
Introducing our new non-saturated (for at least the coming week? 😉) benchmark:
 
✨BigO(Bench)✨ - Can LLMs Generate Code with Controlled Time and Space Complexity?
 
Check out the details below !👇
Baptiste Rozière (@b_roziere) 's Twitter Profile Photo

We released Devstral. It is a 24B model released under the Apache 2.0 license. It the best open model on SWE-Bench verified today. You can check our blog post or test it with OpenHands (from All Hands AI ) following the instructions here: huggingface.co/mistralai/Devs…

Continue (@continuedev) 's Twitter Profile Photo

💻 Agent mode just went local! Today, Mistral AI released Devstral, an Apache-2.0, 24B-parameter model trained for tool calling in real-world software development environments It reached #1 on SWE-bench for open-source models, but more importantly in our real-world testing it

JetBrains (@jetbrains) 's Twitter Profile Photo

Smarter code completion starts with better context. Join the Context Collection Competition by JetBrains and Mistral AI. 🏆 USD 12,000 in prizes 📍 Dedicated workshop co-located with ASE 2025 Ready to build? Start here ⬇️ jb.gg/cccx

Mistral AI (@mistralai) 's Twitter Profile Photo

Announcing Magistral, our first reasoning model designed to excel in domain-specific, transparent, and multilingual reasoning.