Mistral Large 4
What happened
Mistral AI launched a public preview of its new multimodal model, Mistral Large 4 (ML4), also nicknamed “le Chonk.” The 1‑trillion‑parameter model, with 52 billion active parameters, is available via the Mistral Studio preview API, and the full weights will be released at the end of the month. Trained on 3,800 NVIDIA Grace Blackwell GPUs in Europe, ML4 claims state‑of‑the‑art performance on coding, agentic workflows, cybersecurity, finance, law and visual grounding, even surpassing some closed‑source models. It supports over 160 languages and will be deployed in multiple regions, including a Europe‑only instance under EU law.
Why it matters
For ordinary readers, the preview means that a powerful, open‑weight AI model—especially strong in security tasks—will soon be usable and self‑hostable, giving organizations more control over AI tools. Its multilingual and multimodal abilities could broaden access to advanced AI across many industries.
Reality check
An automated, sceptical read of the sourceNotable. The model pushes open‑weight performance forward with a trillion parameters and claims competitive results on specialized benchmarks, but it is still a preview and its full impact remains to be proven.
Is this a big deal?
Get stories like this every morning
The day's AI news in plain English, in your inbox at 8:30 IST. Free.
More from Mistral
Le Chonk brings the brains. Studio puts it to work.Mistral · 5 Nov 2026Le Chonk: Built for Coders, Trusted by Defenders.Mistral · 29 Oct 2026Mistral Advances AI for Engineering with NVIDIAMistral · 21 Oct 2026Voxtral 3Mistral · 20 Oct 2026CERN and Mistral sign a statement of intent to establish a strategic partnershipMistral · 19 Oct 2026This page was collected from mistral.ai and summarised automatically by AIPulsee's pipeline. See how it runs.