Guillaume Lample: Mistral Large 4 launch thread (RL still in flight; 3,800 GB GPUs)

Co-founder thread. 1/n: ML4 is a preview whose RL run 'is still in flight and shows no sign of saturation'; a final version comes with the weights before the end of the month. 2/n (https://x.com/GuillaumeLample/status/2107461902909558833, ~487k views): trained on 3,800 NVIDIA Grace Blackwell GPUs in the Bruyères-le-Châtel cluster built with the Series B; Series C and D clusters are coming online soon.

Today, we are launching a preview of our new model, Mistral Large 4 (ML4), aka le Chonk 🐈. ML4 is a 1T-parameter model with 49B active parameters, trained natively with multimodal capabilities.

It is at the frontier of open models, and by far the strongest open-weight model from the US or Europe. The RL run behind this preview is still in flight and shows no sign of saturation -- we will release a final version before the end of the month along with the weights of the model. 🧵 1/n

Quoting @MistralAI: Meet Mistral Large 4, aka Le Chonk.

• 1T parameters, natively multimodal. 49B active. It is the best open weights model from US or Europe on aggregated benchmarks. • State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses closed frontier models on visual grounding. • Forged in Europe end-to-end and is deployable from Europe via our own Mistral Cloud infrastructure. • Available to all via API today. Working with cybersecurity partners privately.

Open weights release end of October.

Media: https://pbs.twimg.com/media/HT80TJAXIAA_PER.jpg?name=orig https://pbs.twimg.com/media/HT80U4hXgAAFXgi.jpg?name=orig https://pbs.twimg.com/media/HT80WqWXMAAXbjh.jpg?name=orig https://pbs.twimg.com/media/HT80YjcXgAEt9bD.jpg?name=orig

views 199319 · likes 2304 · reposts 263 · replies 113 (at fetch time)

Archived 2026-10-07 via fxtwitter (unofficial).