The French AI lab claims its latest release matches or outperforms even the best models from DeepSeek and Z.ai in coding and cybersecurity.
Mistral has released the weights of Mistral Large 4 on Hugging Face, after three weeks of extra training lifted its coding and cyber defence scores.
Orange Cyberdefense will add Mistral Large 4 to penetration testing, vulnerability checks, and incident response, running it in Europe.
Mistral says its new model can afford organizations a kind of control over its cyber defenses that closed models can't.
Technology
After making Large 4 available in a public preview earlier this month, Mistral is now releasing an improved version to everyone, which it says ranks as the “best open-weight model for coding.” The model offers advancements in cybersecurity-related tasks, like creating patches or reproducing vulnerabilities, and is more likely to refuse “malicious cyber requests” compared to other open models, according to Mistral. [Link: Mistral Large 4 (Le Chonk) is now open weight | Mistral | https://mistral.ai/news/mistral-large-4-oss/ | Mistral]
Technology
French AI startup Mistral says its new trillion-parameter model, Mistral Large 4 — also called Le Chonk — is the best open-weight model developed outside of China. It says it’s particularly strong at coding, cybersecurity, and tasks in finance, engineering, and manufacturing. The model is currently available in preview, with weights dropping at the end of October. [Media: https://twitter.com/MistralAI/status/2107457414387622310?s=20]
Technology
‘We need our own models,’ DeepL’s CEO says, as Mistral’s Mensch warns of ‘leftovers’ if Europe quits frontier AI.