Mistral releases the open weights of Large 4, its 1-trillion-parameter model
Mistral released the weights of Mistral Large 4, nicknamed Le Chonk, on 29 October, after a public preview earlier in the month. It is a natively multimodal mixture-of-experts model with 1 trillion parameters, described as hybrid instruct-and-reasoning, with Low, Medium and High reasoning modes and support for more than 160 languages. The weights are on Hugging Face as Mistral-Large-4-1T-A52B, under a modified MIT licence that allows use, fine-tuning and commercial use, except that companies selling it as a hosted model service need a separate commercial licence.
Mistral reports 41% on Terminal-Bench 4.0, up 13 points from the preview, and 48% on the Artificial Analysis Coding Agent Index, using 47% fewer tokens per task than the leading open-weight model. On security it reports refusing 96.6% of malicious cyber requests on average, and says it is tightening guardrails after seeing some potential for uplift in real attacks; a cybersecurity variant is limited to vetted enterprise partners. API pricing is $0.70 per million input tokens and $2.10 per million output tokens.
An open model at this size gives teams that can't send data to a hosted API a real option. Check the licence before you build a product on it, and test it on your own tasks: the figures are Mistral's.
Drafted with AI from the original source, then checked and edited by Sridhar Mukkandi. Spotted a mistake? Write to hello@tarkika.com and we'll correct it here.