Terminal-first development
Explore a repository, make multi-file changes, run tests and inspect diffs without moving into a separate chat application.
Independent tool overview
Mistral Vibe CLI at a glance
Mistral Vibe CLI is an open-source terminal coding agent that can inspect and edit repositories, run commands, use tools and continue sessions across local, editor and cloud interfaces.
Visit the official Mistral Vibe CLI site ↗
- Best for
- Developers who want an open-source terminal coding agent
- License
- Apache 2.0
- Interfaces
- CLI, web and supported editors
- Model options
- Mistral-hosted, local and OpenAI-compatible
- Installation
- One-line installer, uv or pip
Overview
What Mistral Vibe CLI is
Mistral Vibe CLI provides an interactive coding agent inside the terminal for codebase exploration, multi-file changes, tests, command execution and scripted automation. It launched alongside Devstral 2, but the current Vibe product supports Mistral-hosted models, local models and OpenAI-compatible providers rather than being limited to one model.
The CLI is part of Mistral's wider Vibe Code product, which also runs on the web and in supported editors through Agent Client Protocol. A live session can be moved between the terminal and a remote coding environment with Teleport while retaining its history and approvals.
The software itself is available under Apache 2.0, but hosted inference and higher usage limits depend on a Mistral plan or pay-as-you-go API access. Teams should also review data controls: Free and Pro inputs and outputs may be used for training by default unless the user opts out, while Team and Enterprise data is not used for model training.
Use cases
Who Mistral Vibe CLI is best for
The strongest fit depends on the job you need the product to complete, not the size of its feature list.
Local-model workflows
Run the open-source agent against a local compatible model when data location, cost or offline access matters.
Editor interoperability
Connect the same agent to VS Code, JetBrains, Zed and other Agent Client Protocol clients.
Scripted automation
Invoke Vibe non-interactively for repeatable repository tasks in scripts or CI, with explicit trust and permission configuration.
Capabilities
Core Mistral Vibe CLI features
Repository-aware coding
Reads project files, answers codebase questions, generates code and coordinates edits across multiple files.
Shell and tool execution
Runs approved commands, tests and external tools, with per-tool permissions and boundaries for access outside the working directory.
Flexible model providers
Works with Mistral-hosted models, fully local models or an OpenAI-compatible API configured through profiles.
MCP, skills and subagents
Connects external MCP tools, loads reusable instructions and scripts, and can coordinate specialized subagents for larger tasks.
ACP editor support
Runs through Agent Client Protocol in compatible editors so developers can use Vibe beyond its native terminal interface.
Session portability
Moves an active coding session between the terminal and cloud with Teleport while preserving conversation state and approvals.
Process
How the Mistral Vibe CLI workflow works
Step 1
Install and authenticate
Use Mistral's installer or install the Python package, then sign in or configure a hosted, local or compatible third-party API profile.
Step 2
Review project trust
Open the repository, inspect any project-level .vibe configuration and trust the folder only after reviewing its agents, MCP servers and permissions.
Step 3
Give a bounded task
Describe the desired outcome, relevant files and acceptance checks; reference files directly when the agent needs a precise starting point.
Step 4
Approve and inspect changes
Review proposed shell commands, file access and highlighted diffs instead of granting broad automatic permissions on sensitive machines.
Step 5
Verify or hand off
Run tests locally, continue interactively or teleport a longer task to the remote environment and review the resulting change before merging.
Cost
Mistral Vibe CLI pricing and free plan
The Vibe CLI code is open source. Mistral-hosted usage is available with limited Free access, the $14.99 per month Pro plan, the $24.99 per user per month Team plan or custom Enterprise terms. Monthly API credits are shared across Vibe Code, Studio and API usage, and additional usage can be pay-as-you-go.
Free
$0/month
For testing Vibe and limited coding sessions.
- Limited coding sessions
- $10 per month in included API credits
- Access to Mistral Studio
- Local and compatible-provider use can have separate infrastructure costs
Pro
$14.99/month
For individual developers who want higher limits and all-day coding.
- Full access to Vibe coding across CLI, editor and web
- $30 per month in included API credits
- Higher task and usage limits subject to fair-use rules
- Chat and email support
Team
$24.99/user/month
For collaborative workspaces with organizational controls.
- Secure shared workspace
- Up to 30GB storage per user
- Domain verification and data export
- Inputs and outputs are not used to train Mistral models
Enterprise
Custom
For private deployment, governance and model customization.
- Private cloud, on-premises or Mistral Cloud deployment options
- Custom models, agents and workflows
- Audit logs and SAML SSO
- Dedicated solutioning and support
Pricing checked . Check current pricing at the source ↗
Assessment
Mistral Vibe CLI strengths and limitations
Where it stands out
- Open-source agent with a permissive Apache 2.0 license
- Supports Mistral-hosted, local and OpenAI-compatible model endpoints
- Available across terminal, editor and remote web workflows
- Layered folder, tool and out-of-directory permission controls
- Active development with MCP, skills, subagents, programmatic mode and Teleport
What to consider
- Free and Pro hosted inputs and outputs may be used for model training by default unless the user opts out
- The CLI can edit files and execute shell commands, so trusting a repository or broad auto-approval rules can expose local secrets and credentials
- Hosted usage is subject to plan limits, shared monthly allowance and model-specific pay-as-you-go costs
- Windows is supported through manual installation, but the open-source project says its primary target remains Unix environments
- Model quality and behavior depend on the selected provider; an open-source harness does not make hosted inference free
Compare
Mistral Vibe CLI alternatives
The right alternative depends on the specific output, workflow, controls and budget your project requires.
Coding
Codex
Choose Codex for OpenAI's actively maintained local-and-cloud coding agent with strong parallel task orchestration.
Explore Codex →Coding
Claude Code
Choose Claude Code for Anthropic's terminal agent and deep integration with the Claude model family.
Explore Claude Code →Coding
Gemini CLI
Consider Gemini CLI for an open-source terminal agent with Google's Gemini models and a generous free entry point.
Explore Gemini CLI →Coding
SWE-1.7
Consider SWE-1.7 inside Devin when you want Cognition's current fast software-engineering model and managed agent environment.
Explore SWE-1.7 →Questions
Mistral Vibe CLI FAQs
What is Mistral Vibe CLI?
It is Mistral's open-source terminal coding agent. It can read a project, edit files, run approved commands, call tools and work interactively or programmatically.
Is Mistral Vibe CLI free?
The CLI source is free under Apache 2.0, and Mistral offers limited hosted coding on its Free plan. Hosted inference beyond included usage, third-party APIs and local compute can create separate costs.
Does Vibe CLI require a Mistral model?
No. The current setup supports Mistral-hosted models, local models and OpenAI-compatible API providers configured through profiles.
Can Vibe CLI run offline?
Yes, when configured with a compatible model running locally. Features that require Mistral services, remote sessions or online connectors will still need network access.
Does Mistral train on Vibe CLI data?
For Free, Pro and Education accounts, Vibe inputs and outputs are used for training by default unless the user opts out. Team and Enterprise inputs and outputs are not used for model training.
Is Vibe CLI safe to run on a work repository?
It provides trusted-folder and per-tool permission controls, but it can read files, edit code and execute commands. Review repository-level .vibe configuration and avoid broad permissions on machines containing production credentials.
Bottom line
Our Mistral Vibe CLI verdict
Mistral Vibe CLI is a compelling option for developers who value an open-source agent harness, provider choice and movement between terminal, editor and cloud. Its flexibility also increases operational responsibility: review project configuration, keep permissions narrow, choose the data plan deliberately and evaluate the selected model on your own codebase.
Visit Mistral Vibe CLI website ↗Similar Coding Tools
View all Coding AI tools →
Codex Max
Codex Max - OpenAI's new frontier agentic coding model
%20(1).png)
Devstral 2
Devstral 2 - Mistral's next-gen family of coding-specific models
Antigravity
Antigravity - Google's new agentic development platform to build anything, orchestrate agents, and run parallel tasks across workspaces.
Devin Desktop (formerly Windsurf)
Codeium Windsurf: Provides ai-powered code suggestions, completions, and debugging tools for developers.
AI training for the future of work.
Get access to all our AI courses, hundreds of real-world AI use cases, live expert-led workshops, an exclusive network of AI early adopters, and more.
AI Courses
Get unlimited access to all of our current & upcoming industry-specific AI courses for the duration of your subscription.
Daily Guides
To keep up with the rapid pace of AI, our team publishes AI implementation guides daily. Our library contains 300+ practical use cases to automate real-world work.
Workshops
Join weekly, live, interactive sessions with industry leaders who are at the forefront of AI for hands-on implementation guidance and exclusive insights.
Community
Network with an exclusive community of AI-first professionals who are working smarter with AI. Learn how early adopters are using AI in their work and businesses.