Qwen Code
An open-source terminal coding agent from the Qwen team that works with Qwen, OpenAI, Anthropic, Gemini and local models
Install
npm install -g @qwen-code/qwen-code@latestRequires Node.js 22+.
This is third-party code. Review the repository files before installing.
What it does
Qwen Code starts with the qwen command inside a project and lets you talk to the codebase: it explains the structure, edits files and runs commands. It supports subagents, hooks, skills, MCP, plan mode and git worktrees. The model provider can be switched at runtime, including Ollama or vLLM. Besides the terminal there is a desktop app, a web UI via qwen serve, plugins for VS Code, Zed and JetBrains, and a headless qwen -p mode for scripts and CI.
Who it is for. For developers who want an open alternative to Claude Code with a free choice of model, local ones included.
Good fit when
- You need a terminal agent not tied to one model vendor
- You want to work with a local model via Ollama or vLLM
- You need to run an agent in CI in headless mode
Not a fit when
- Your team has standardized on another agent and has no plans to switch
- You have neither cloud API access nor hardware for a local model
Example request
Explain how this repository is organized and show me where to startLimitations
It needs a model provider key or a local model server. Access to cloud APIs from Russia depends on the provider: OpenAI, Anthropic and Gemini do not work without a VPN, local models do. The npm install requires Node.js 22+. The web UI and daemon mode are marked experimental.
How to disable. Remove the package with npm uninstall -g @qwen-code/qwen-code or brew uninstall qwen-code; settings live in ~/.qwen.
Security check
- The agent runs shell commands and modifies project files
- Project code is sent to the chosen model provider
README in short
The README covers installation via script, npm or Homebrew, and the first run with /auth. It lists the interfaces: terminal, desktop, web, editors and chat app channels. It includes a feature table compared with Claude Code and SWE-bench Verified results. SDKs for TypeScript, Python and Java are mentioned. Apache-2.0 licensed.
FAQ
Is it a Gemini CLI fork?
It started from Gemini CLI, but since version 0.1 it is developed independently and no longer syncs with upstream.
Can it run without cloud models?
Yes, the README lists local model support via Ollama and vLLM.
Related
Open-source personal AI assistant on your own machine: answers in Telegram, Slack, Discord and WhatsApp, extended with skills and plugins
A self-improving agent from Nous Research with a TUI, messaging gateway, cron jobs and skills it writes itself
Anthropic's coding agent for the terminal, IDE and GitHub: understands your codebase, runs tasks and handles git
OpenAI's coding agent that runs locally in your terminal, with IDE and desktop versions