Qwen Code
FreeAlibaba's terminal coding agent, optimized for Qwen models.
About Qwen Code
Qwen Code is an open-source AI coding agent that operates directly within the terminal, offering a comprehensive set of agentic capabilities out of the box. It features Auto-Memory, Auto-Skills, SubAgents, Agent Teams, and MCP, enabling dynamic workflows with zero setup. The tool supports multiple protocols including OpenAI, Anthropic, Gemini, and Qwen APIs, as well as local models via Ollama or vLLM. Beyond the terminal, Qwen Code integrates with IDEs (VS Code, Zed, JetBrains), offers a desktop app, daemon mode, SDKs (TypeScript, Python, Java), and IM bots (Telegram, DingTalk, WeChat, Feishu). The framework and underlying Qwen models are both open-source, ensuring no vendor lock-in. Notably, Qwen Code actively iterates on itself by using its own agent and models to file issues, submit PRs, review code, and run tests.
Key Features
Pros & Cons
- Completely open-source and free to use
- Supports a wide range of LLM providers without vendor lock-in
- Rich agentic capabilities like memory and subagents out of the box
- Versatile deployment options (terminal, IDE, desktop, daemon, bot)
- Actively maintained and self-improving via automated PRs and reviews
- Requires Node.js 22+ or npm for installation via NPM
- Terminal-based primary interface may have a learning curve for some users
- Relatively new project with an evolving ecosystem