qwen-code
by
qwenlm

Description: An open-source AI agent that lives in your terminal.

View qwenlm/qwen-code on GitHub ↗

Summary Information

Updated 2 hours ago
Added to GitGenius on February 18th, 2026
Created on June 26th, 2025
Open Issues/Pull Requests: 570 (+0)
Number of forks: 1,689
Total Stargazers: 19,360 (+4)
Total Subscribers: 106 (+0)
Detailed Description

Qwen Code is an open-source AI agent designed to reside within your terminal, providing developers with a powerful tool for understanding, automating, and accelerating their coding workflows. Built upon the foundation of the Qwen3-Coder model, this agent offers a terminal-first experience, with optional integrations for popular IDEs like VS Code, Zed, and JetBrains IDEs. Its primary purpose is to assist developers in navigating large codebases, automating repetitive tasks, and ultimately, shipping code faster.

The core functionality of Qwen Code revolves around its ability to interact with code through natural language prompts. Users can ask questions about the project's structure, request code refactoring, generate unit tests, and perform a wide range of other coding-related tasks directly within the terminal. The agent leverages the power of the Qwen3-Coder model to understand these prompts and generate relevant responses, code snippets, or suggestions.

A key feature of Qwen Code is its flexibility in authentication and model selection. It supports multiple API protocols, including OpenAI, Anthropic, and Google Gemini-compatible APIs. Users can choose to authenticate using Qwen OAuth, which provides 1,000 free requests per day, or utilize API keys for greater control over the model and provider. The configuration of models and providers is managed through a `settings.json` file, allowing users to define available models, API keys, and default settings in a centralized location. This configuration supports various providers like OpenAI, Anthropic, and Gemini, as well as Alibaba Cloud Bailian.

Installation is straightforward, with quick install scripts available for Linux/macOS and Windows. Alternatively, users can install Qwen Code via NPM or Homebrew. Once installed, the agent can be launched in interactive mode by simply typing `qwen` in the terminal. Headless mode is also supported, enabling the agent to be used in scripts and CI/CD pipelines. Furthermore, the project offers a TypeScript SDK, allowing developers to build upon Qwen Code's capabilities.

The interactive mode provides a user-friendly terminal UI with commands like `/help`, `/clear`, and `/auth`. Keyboard shortcuts are also available for efficient navigation and control. The agent's functionality extends to IDE integration, enhancing the coding experience within popular editors.

The project emphasizes its open-source nature, allowing for community contributions and co-evolution with the Qwen3-Coder model. The documentation provides comprehensive guides on installation, authentication, usage, and configuration. The project also includes benchmark results demonstrating its performance in code-related tasks.

Qwen Code aims to be a comprehensive tool for developers, offering a blend of AI-powered assistance and terminal-based convenience. It is designed to streamline coding workflows, improve code understanding, and boost developer productivity. The project's focus on open-source development, flexible authentication, and IDE integration makes it a versatile and accessible tool for a wide range of developers. The project also provides resources for troubleshooting and connecting with the community, ensuring a supportive environment for users.

qwen-code
by
qwenlmqwenlm/qwen-code

Repository Details

Fetching additional details & charts...