Mistral rebrands Le Chat to Vibe, launches autonomous coding agent and work automation platform

Mistral AI has rebranded its Le Chat conversational platform as Vibe, launching two specialized agent modes designed for complex work automation and autonomous coding. The platform, announced May 28, 2025, carries over all existing user conversations, settings, and plans from Le Chat.

Work Mode and Code Mode

Work Mode operates as an agent for multi-step business tasks, connecting to Google Workspace, Outlook, SharePoint, Slack, GitHub, and custom connectors. The agent creates a plan, requests user approval before execution, and streams progress as it works. According to Mistral AI, it handles enterprise knowledge search, structured data analysis, document synthesis, and multi-step task scheduling on daily, weekly, or monthly cadences.

Code Mode launches remote coding agents from a dedicated web interface at code.mistral.ai. Sessions run in isolated sandboxes and persist even when the user's machine is off. The agent handles feature development, bug fixes, refactoring, and ships pull requests. Mistral AI claims the agent works across the entire codebase with context from GitHub, GitLab, Jira, and Linear.

VS Code extension and CLI updates

Mistral released a new VS Code extension that embeds the Vibe coding agent in a side panel. The extension reads, edits, and executes commands across the entire project. Open files attach automatically, and users can reference specific line ranges or pull context from other directories using @ mentions.

The Vibe CLI receives updates including custom skills as slash commands, editable agent plans before execution, session-scoped permissions, and a /teleport command that moves live sessions between terminal and cloud while preserving history and approvals.

Pricing and availability

Vibe operates on four pricing tiers:

• Free: Basic queries and simple tasks
• Pro: $14.99/month for complex tasks, deeper reasoning, and full-day coding
• Team: $24.99/user/month with shared workspace and admin controls
• Enterprise: Custom deployments with model training and dedicated support

The platform runs on flagship Mistral models optimized for reasoning, agentic tasks, tool calls, and coding, though Mistral did not specify which model versions power Vibe.

What this means

Mistral AI is positioning Vibe as a unified agent platform competing directly with GitHub Copilot Workspace, Anthropic's Claude with computer use, and OpenAI's ChatGPT Enterprise. The $14.99 Pro tier undercuts GitHub Copilot's $10/month by offering both coding and work automation, though it remains unclear how Vibe's code quality compares to established coding assistants. The ability to run persistent coding sessions in cloud sandboxes and move them between terminal and web distinguishes Vibe from editor-bound competitors, but enterprise adoption will depend on how well the agent handles production codebases and whether its multi-app integrations can reliably complete complex workflows without human intervention.