qwen-code

qwen-code

Open-source AI coding agent that works in your terminal, IDE plugins, a desktop app, a web page and chat apps, with sub agents, MCP and many model providers. For developers who code with AI.

Made by QwenLM on GitHub. These notes are Mindket's, written from its README; the project's own files and license are in its repository.

About

What it is: Qwen Code is an open-source AI coding agent. It works in your terminal first, and also in editors, a desktop app, a browser and chat apps. Its built-in features include memory, skills, sub agents, agent teams and MCP.

The framework and the Qwen models are both open source, and the project says there is no vendor lock-in.

Who it suits: It suits developers who want an AI helper for their code and want a choice of model, hosted or local.

Industry | Who uses it | How they use it | Example
Software | Developers who live in the terminal | Hand coding tasks to an agent from the command line | Asking the agent to look into a bug in their project
Teams | Groups that talk in chat apps | Reach the agent from Telegram, DingTalk, WeChat or Feishu | Posting a coding question in the team's chat
AI tooling | Developers building on agents | Use the SDKs to build on the agent | Wiring the agent into an internal tool

Anyone who does not write code, or who prefers a single hosted assistant with no choice of model, can skip it.

Where to use it:

• Your terminal
• IDE plugins and a desktop app
• A web UI
• SDKs, for building on the agent
• Chat integrations for Telegram, DingTalk, WeChat and Feishu

Use case | Who | What they get back
Work on code from the terminal | Developers | An AI agent that handles coding tasks
Use it inside an editor | IDE users | The same agent through a plugin
Ask from a chat app | Teams | The agent reachable from their chat
Build on the agent | Developers | SDKs to build on

How to set it up:

• Git, to copy the code from its GitHub page
• Access to a model: an OpenAI, Anthropic, Gemini or Qwen API, another provider, or a local model through Ollama or vLLM

1. Copy the project's code from its GitHub page with git.
2. Open its README and follow the setup steps there.
3. Connect the model provider you want to use.

The kit holds notes and a link to the code, and nothing in it runs on its own.

For newcomers and experts: If you are new, a coding agent is an AI helper that works on your code when you describe the job in plain words. Qwen Code does this from your terminal. A first small thing to try is connecting one model and asking it about a small project.

If you know the field, the interesting parts are the protocol support and the agent features. It speaks the OpenAI, Anthropic, Gemini and Qwen APIs, takes third-party providers and local models, and lets you switch model while it runs. On top of that sit sub agents, agent teams, MCP, memory and skills.

You ask for a coding task in a terminal, editor or chat
The agent sends it to the model you picked
Sub agents, skills and MCP tools help where needed
The result comes back where you asked

Good to know:

• Self-use: the project says its own agent and models help run its development, including testing and code review
• Languages: the README is also offered in several other languages
• License: Apache-2.0
• Cost to run: no price named. You need access to a model, either an API from a provider or a local model.
• Data it reaches: none named. The model you pick decides where your prompts go.

On GitHub, 13 November 2026

• Stars: 29,845 (151 that day)
• Forks: 3,399
• Watching: 146
• Open issues and pull requests: 2,225
• Language: TypeScript
• License: Apache-2.0
• Topics: agentic, ai, ai-agent, ai-coding, cli, coding-agent, developer-tools, llm, mcp, qwen, terminal, typescript
• Trending: #1 repository of the day on GitHub trending (13 November 2026)

Links

• Code: https://github.com/QwenLM/qwen-code
• Website: https://qwenlm.github.io/qwen-code-docs/en/users/overview
• Releases: https://github.com/QwenLM/qwen-code/releases
• Issues: https://github.com/QwenLM/qwen-code/issues
• Star history: https://www.star-history.com/#QwenLM/qwen-code&Date

Get the code

git clone https://github.com/QwenLM/qwen-code.git

Then follow the setup in its README. Nothing in this kit runs by itself.