Overview
Agent Zero is an open-source agent framework that gives an AI agent a complete computer to work with. A single Docker container ships a full Linux system, including an XFCE desktop and a built-in browser, so the agent can use the terminal, run code, and drive real desktop software to finish a task.
It is aimed at developers and tinkerers who want a general-purpose agent that operates a real machine rather than calling a fixed set of APIs. You configure your own LLM provider in the Web UI and then hand the agent concrete tasks, watching each action as it happens.
Within the computer-use category, Agent Zero stands out by exposing an actual Linux desktop and a native browser side by side. The agent can manage files visually, open GUI tools that have no API, and read or edit web pages, while you can step in at any moment because you share the same desktop.
What it does
- Ships a full Linux system with an XFCE desktop inside one Docker container
- Native browser with an Annotate mode to inspect, change, lift, or comment on page elements
- Markdown editor and LibreOffice (Writer, Calc, Impress) for live cowork on documents
- Plugin Hub with 100+ community plugins, installable with a click from the Web UI
- Extensible through custom prompts, custom tools, MCP servers, and A2A connectors
- A0 CLI Connector bridges a running instance to your host machine and your own browser
Getting started
Install with the one-line script for your OS, or run the Docker image directly if you already have Docker, then open the Web UI to configure your LLM provider.
Install on macOS or Linux
Run the official install script in your terminal.
curl -fsSL https://bash.agent-zero.ai | bashInstall on Windows
Run the install script from PowerShell.
irm https://ps.agent-zero.ai | iexOr run the Docker image directly
If Docker is already installed, start Agent Zero with one command. The volume keeps your data between runs.
docker run -p 80:80 -v a0_usr:/a0/usr agent0ai/agent-zeroOpen the Web UI and start a task
Open the Web UI in your browser, configure your LLM provider, then give the agent a concrete task to start with.
Commands and code are distilled from the project's own documentation — always check the official repo for the latest.
When to use it
- Drive GUI desktop software that has no API, such as modeling a 3D object in Blender
- Browse and edit web pages, then re-implement a component you like in your own stack
- Cowork on plans, TODOs, or notes in a live Markdown or LibreOffice document
- Automate multi-step file and terminal work on a real Linux machine you can watch
How Agent Zero compares
Agent Zero alongside other open-source computer & browser use tools AI/TLDR tracks, ranked by GitHub stars.
| Tool | Stars | What it does |
|---|---|---|
| Browser Use | ★ 105k | A Python library that lets agents control a real browser to read pages and complete tasks online from natural-language instructions. |
| Open Interpreter | ★ 65.9k | A lightweight coding agent that runs code on your own computer to carry out tasks from natural-language requests. |
| UI-TARS Desktop | ★ 38k | ByteDance's multimodal agent stack and desktop app that controls a computer's graphical interface using vision-language models. |
| Stagehand | ★ 23.5k | A TypeScript browser-automation SDK from Browserbase that mixes natural-language actions with normal code for reliable web agents. |
| Skyvern | ★ 22.4k | A tool that uses language models and computer vision to automate browser workflows without writing custom code for each website. |
| Cua | ★ 19.9k | Infrastructure for computer-use agents, providing sandboxes and SDKs that let agents control full desktops on macOS, Linux, and Windows. |
| Agent Zero | ★ 18.4k | A personal AI agent with a full Linux system, desktop, and browser as its tools |
| Browser Use Web UI | ★ 16.2k | A user-friendly web UI, built on Gradio, that lets you control browser-use AI agents, pick from many LLMs, and use your own browser. |