Qwen Code running in a terminal saves pages into your Showspace, and it can use all twenty of the app's commands.
Qwen Code is an open-source coding agent from Alibaba's Qwen team. You run it inside a folder and it edits the files there, so a dashboard or a project page is as natural a job as a code change. It works with Qwen, OpenAI, Anthropic and Gemini APIs, and with local models through Ollama or vLLM.
On a Mac it runs in a terminal, installed with Homebrew, npm or the install script. Qwen Code Desktop, the official app for Apple Silicon and Intel Macs, offers the same capabilities in a window.
brew install qwen-code.Here is a starter prompt. How well it works depends on how your AI is set up, so adjust it to fit yours.
Build an interactive API reference for this project: a searchable list of endpoints, tabs for each endpoint's request and response, and a copy button on every example. Save it as a single self-contained HTML file in my Showspace, then open it in lite.computer.
qwen channel start you can send it tasks from Telegram, DingTalk, WeChat or Feishu on your phone. The work happens on your Mac, so pages land in your Showspace.Yes. Qwen Code running on your Mac saves HTML pages into your Showspace, and lite.computer shows each one as it lands. Once it is connected to the app, it can open pages and arrange windows there as well.
The software is free and open source. Models come through an Alibaba Cloud Coding Plan, a Token Plan or an API key from any supported provider.
Choosing between agents? See Which AI works with lite.computer.
Last checked September 14, 2026.
Still stuck? Write to support@lite.computer.