Mistral is making a bold move into the crowded world of AI coding assistants with its new tool, new tool for professional developers.
Designed to appeal to professional developers, Mistral Code is a full-featured coding companion now in private beta for JetBrains and VS Code environments.
Mistral Code sets itself apart by combining its own AI models with open source technology, offering fast code completions, multi-step refactoring, and the ability to work entirely on local machines or private networks.
The company has packed the assistant with a host of features for both individual developers and IT teams.
More than eighty programming languages are supported, and users can expect everything from code search to chat guidance, plus support for many third-party plugins.
In a nod to enterprise concerns, Mistral allows customers to train its models on their own code bases, giving teams more control and better accuracy for their unique needs.
The client’s admin console also delivers in-depth controls for managing seats, checking usage analytics, and overseeing security within organizations.
Early Production Use and Customer Adoption
Early adopters include heavyweights such as French rail giant SNCF, international bank Abanca, and consulting firm Capgemini.
These companies are already using Mistral Code in their development pipelines.
Mistral’s models powering the client include Codestral for code autocomplete, Codestral Embed for searching and retrieving snippets, Devstral for automating more complex coding tasks, and Mistral Medium for chat-based assistance.
Customers are also able to fine tune or further train these models using their own data, or create more compact variants for lighter applications.
Developers working with the assistant can get help analyzing terminal outputs, navigating source files, and tackling issues — all directly inside their regular development tools.
The pricing and availability for Mistral Code’s wider public launch has yet to be announced, but the company is signaling big ambitions in the enterprise space.
Founded only last year, Mistral has built up a hefty €1.1 billion in investment and is quickly rolling out new products, including a corporate chatbot and a suite of AI models like Codestral and Devstral.
The startup says it plans to contribute its own work back to the open source community through ongoing updates to the Continue project — the same platform that Mistral Code is built upon.
With surging interest in adoption of AI programming help — more than seventy percent of developers are either using or planning to use these tools, according to recent research — Mistral’s bet on customizable, secure, and local-ready coding AI seems well timed.
