Qwen Code
An open source coding assistant for the terminal and desktop. It runs on macOS, Windows and Linux and works with Ollama, vLLM and cloud model APIs.
Qwen Code is an Apache 2.0 licensed AI coding agent for developers who want help working through a codebase, changing code and checking the result. It runs on macOS, Windows and Linux, with a terminal interface and a desktop app. It builds on Google Gemini CLI and has developed into an agent that can use Qwen models alongside other model providers.
Model choice is a practical reason to consider it. Qwen Code works with locally hosted models through Ollama or vLLM, and with Qwen, OpenAI, Anthropic and Gemini APIs. You can switch providers during a session. The agent runs on your machine, while model requests use the backend you choose; a cloud API involves its provider.
Beyond chat, Qwen Code can review code, investigate bugs, plan work and carry out coding tasks. Its agent features include memory, skills, subagents and agent teams. MCP support lets it use external tools, while sandboxing and Git worktrees give it ways to work on changes in an isolated environment. It can also use a computer desktop for automation.
Developers can use Qwen Code in VS Code, JetBrains or Zed, as well as through a browser interface. Headless operation and TypeScript, Python and Java SDKs cover scripted workflows. For teams comparing coding assistants across models, Agent Arena runs different models on the same task. Chat connections include Telegram, DingTalk, WeChat and Feishu.
