Le Chat, Mistral AI’s chatbot, is getting a major update with five new features, including a deep research mode, voice control, and a revamped image editing tool. The rollout uses several specialized models, like Voxtral for speech and Magistral for logic and reasoning.
The highlight is Deep Research Mode, a tool already seen in products like ChatGPT, Gemini, Grok, and Claude. With this feature, Le Chat can break down complicated questions into smaller parts, search through multiple sources, and produce a structured report with references.
As with other deep research tools, these reports are bound to contain errors, and the quality of sources can be inconsistent. Still, the feature is useful for experts who want to quickly find sources or get an overview of a new topic in their field—situations where you don’t need to fact-check every detail, just the main points.
Voice input and image editing
Le Chat now supports direct voice input with the Voxtral model. The voice mode is built for low latency and natural speech, making it possible to brainstorm or take notes by simply talking to Le Chat, the company says.
For complex reasoning tasks, the new “Think Mode” uses Mistral’s Magistral model, though benchmark. The image editing tool, developed with Black Forest Labs, lets users modify generated images through text commands—removing objects or changing scenes while preserving key elements. The feature likely uses Black Forest Labs’ Kontext models.
A new projects feature organizes conversations, files, and settings into workspaces, automatically saving tools used and allowing file uploads. Mistral says this works particularly well for long-term projects.
All features are live now at chat.mistral.ai and in Le Chat’s iOS and Android apps.
