{"slug": "introducing-mistral-large-4-le-chonk", "title": "Introducing Mistral Large 4: Le chonk", "summary": "Mistral released Mistral Large 4, a preview available through its API, with open weights promised for the end of this month, according to Mistral. The model supports only two reasoning levels, \"none\" and \"high\", and scores 38 on Artificial Analysis, just behind the 552B-parameter DeepSeek 4.1 Flash and far above Mistral Large 3's score of 9. In a pelican SVG test, the \"high\" reasoning setting produced a better result using 2,717 output tokens versus 3,275 for \"none\".", "body_md": "[Introducing Mistral Large 4: Le chonk](https://mistral.ai/news/mistral-large-4/)\n\nThe preview is available via their API. They promise to release the open weights model at the \"end of this month\".\n\nThe model only supports two reasoning levels - \"none\" and \"high\" - via the Mistral API. Here are [both pelicans](https://tools.simonwillison.net/markdown-svg-renderer?url=https%3A%2F%2Fgist.github.com%2Fsimonw%2F0db8b8196e57711b352f6c5857cf5b35) - the \"high\" one looks better, though surprisingly it only used 2,717 output tokens compared to \"none\" which used 3,275:\n\nOn Artificial Analysis [it scores 38](https://artificialanalysis.ai/models/mistral-large-4), just behind DeepSeek 4.1 Flash, which is a 552B model. It's a *huge* improvement on last December's Mistral Large 3, which drew [this terrible pelican](https://tools.simonwillison.net/markdown-svg-renderer?url=https%3A%2F%2Fgist.github.com%2Fsimonw%2F0df5e656291d5a7a1bf012fabc9edc3f#response-1) and [scored 9 on AA](https://artificialanalysis.ai/models/mistral-large-3).\n\nIt's certainly not a Fable-class model, but it's great to see Mistral put out a model that's back to being maybe about 6 months behind the frontier.\n\nVia [Hacker News](https://news.ycombinator.com/item?id=49977979)\n\nTags: [ai](https://simonwillison.net/tags/ai), [generative-ai](https://simonwillison.net/tags/generative-ai), [llms](https://simonwillison.net/tags/llms), [mistral](https://simonwillison.net/tags/mistral), [pelican-riding-a-bicycle](https://simonwillison.net/tags/pelican-riding-a-bicycle), [llm-release](https://simonwillison.net/tags/llm-release)", "url": "https://wpnews.pro/news/introducing-mistral-large-4-le-chonk", "canonical_source": "https://simonwillison.net/2026/Oct/6/le-chonk/", "published_at": "2026-10-06 20:18:19+00:00", "updated_at": "2026-10-07 15:17:27.640290+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "generative-ai", "ai-products"], "entities": ["Mistral", "Mistral Large 4", "Mistral Large 3", "DeepSeek 4.1 Flash", "Artificial Analysis", "Hacker News"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/introducing-mistral-large-4-le-chonk", "markdown": "https://wpnews.pro/news/introducing-mistral-large-4-le-chonk.md", "text": "https://wpnews.pro/news/introducing-mistral-large-4-le-chonk.txt", "jsonld": "https://wpnews.pro/news/introducing-mistral-large-4-le-chonk.jsonld"}}