Mistral Vibe
Mistral Vibe is Mistral's open-source CLI coding assistant. Rechecked 2026-08-25: github.com/mistralai/mistral-vibe had 4,872 stars, license Apache-2.0. The CLI itself is $0. You need a Mistral API key from console.mistral.ai. We did not re-audit every Model Studio row, so this page has no invented dollar seat. Official support targets UNIX; Windows works but is not the first-party target.
Compare Claude Code if you wanted Anthropic's CLI, Qwen Code if you wanted another lab-backed open CLI, and Cline if you wanted a $0 individual agent plus BYOK.
Key Features
- Conversational CLI: read, write, patch, grep, bash, todos,
ask_user_question, andtasksubagents. - Built-in agents:
ask,plan,accept-edits(default),auto-approve.--agent auto-approveor--yoloapproves every tool. - Skills + MCP: Agent Skills spec, custom slash commands, and MCP servers (
http,streamable-http,stdio). - Programmatic mode:
vibe --prompt "..."with--max-turns,--max-price,--max-tokens, and JSON output. - Install:
curl -LsSf https://mistral.ai/vibe/install.sh | bash, oruv tool install mistral-vibe, orpip install mistral-vibe.
Limitation: the default agent auto-approves file edits. --yolo auto-approves everything. Use that only in a trusted folder. The trust-folder prompt exists for a reason.
Use Cases
- People already on Mistral API who want a first-party TUI instead of wrapping the key in Cline.
- People who want
--agent planfor read-only exploration before edits. - People scripting headless runs with
--max-priceas a hard stop.
Pricing
CLI is free. Model access on 2026-08-25.
| Path | Price | Notes from first-party pages |
|---|---|---|
| Mistral Vibe CLI | $0 | Apache-2.0. 4,872 stars on the check date. |
| Mistral API key | Usage | Get a key at console.mistral.ai. Recheck the live Model Studio table. Do not invent a $20 seat. |
--max-price | Session cap | Programmatic runs can stop at a dollar budget you set. |
If a leftover blog prints a Vibe Pro $20 row, treat this table as the source.
Getting Started
- Install with
curl -LsSf https://mistral.ai/vibe/install.sh | bash, oruv tool install mistral-vibe. cdinto a repo and runvibe. Paste a Mistral API key on first launch (saved to~/.vibe/.env).- Stay on
accept-editsuntil you trust the repo. Use--agent planto explore. - Do not pass
--yoloor--agent auto-approveoutside a disposable sandbox.
First-party start: github.com/mistralai/mistral-vibe.
Frequently Asked Questions
Is the CLI still free?
Yes. The Apache-2.0 CLI is $0. Models are billed by Mistral.
Official on Windows?
It runs. The README says UNIX is the supported target.
Same as Le Chat?
No. Le Chat is Mistral's consumer chat. This page is the vibe coding CLI.
Alternatives
- Claude Code: Anthropic first-party CLI.
- Qwen Code: Alibaba open CLI plus ModelStudio.
- Cline: $0 individual agent plus BYOK.
Tips
- Quote 4,872 stars and Apache-2.0 from the GitHub check.
- Do not invent a Vibe dollar seat.
- Set
--max-priceon unattended--promptruns.
Conclusion
Mistral Vibe is a 4,872-star Apache-2.0 terminal agent that bills through a Mistral API key, not a leftover $20 seat. Start at github.com/mistralai/mistral-vibe, then decide whether Claude Code already covers the workflow you need.
Comments
No comments yet. Be the first to comment!
Related Tools
Autohand Code
www.autohand.ai/code
Self-evolving terminal agent. Rechecked: Apache-2.0 with ARR cap, 178 stars. CLI $0 plus BYOK. Enterprise is sales, not a public $20 seat.
MiMo Code
mimo.xiaomi.com/coder
Xiaomi terminal agent. Rechecked: MIT, 12,863 stars. OpenCode fork. CLI $0; Xiaomi OAuth or BYOK. Not a published dollar seat.
OpenClaude
openclaude.gitlawb.com
Open terminal agent. Rechecked: 30,818 stars. MIT for contributor changes; derived from Claude Code. CLI $0 plus BYOK.
Related Insights
What Locks Codex Is the Picker, Not the Models
You already pay for OpenCode Go, Grok, and Z.ai, but the Codex picker still shows mostly GPT. The community project codex-router does not teach another install ritual. It puts subscriptions you already bought back into the selector, keeps keys on the machine, and leaves native GPT alone.
Hook OpenCode Go into Codex on Windows. Do Not Open a Second Toolkit.
A ChatGPT-signed Codex desktop app still shows mostly GPT in the picker. On Windows, enable only OpenCode Go and the Grok, GLM, Kimi, DeepSeek, and MiniMax models you already pay for appear in the same selector. Keys stay local. Native GPT stays put.
Seven AI Coding CLIs, Six Months: No Matter How Strong the Model, Work Needs Supervision
Claude Code, Codex, opencode, pi, omp and DeepSeek Harness all have personalities. After six months of deep use I run a division of labor: pi for the fastest cheapest reviews, omp for complex PRs, DeepSeek Harness on V4 Flash for high-frequency low-cost review, and Claude Code, Qoder and Cursor for writing. No matter how strong the model, work needs supervision — ideally from an independent third party.