Mistral, the French AI company, releases its Large 4 model for download, positioning it as a tool for companies that want more control over how AI is used for cybersecurity and coding. Both outlets say the model supports “agentic” coding tasks and cyber-defence capabilities, and is designed to run using an organization’s own infrastructure.
In context, the release is framed around the growing interest in “open” models that can be deployed outside closed, proprietary systems. Euronews emphasizes that Large 4 can be run on a company’s own computers and servers, linking that control to cybersecurity. Fortune focuses more on performance claims, describing Mistral’s assessment of Large 4 as a top open model for agentic coding and cyber defense, and highlights the business pitch to firms concerned about reliance on closed AI.
