Practical overview
When it fits
Use OpenBrowser when a task needs browser navigation and interaction, especially when an LLM-generated sequence of Python actions is useful for completing multi-step workflows or extracting page data.
Who it is for
- Developers building browser-based automation with Python.
- Teams automating multi-step web interactions, such as product research, form filling or booking flows.
- AI assistant users who want to connect browser automation through the documented MCP server or Claude Code plugin.
Documented uses
- Navigate through websites and interact with page elements.
- Extract information from web pages, including product details.
- Fill out web forms through generated browser actions.
- Record and replay browser workflows, with documented export to notebooks or crawler code.
Requirements & limits
The project is open source under the MIT license. The README documents installation through pip, uv, uvx, pipx, Homebrew and platform-specific installers. Its hosted cloud dashboard for saved cookies and scheduled tasks is described as available by waitlist for early access.
- The documented agent setup uses Python 3.12 or later, an LLM provider credential and a Chromium-based browser.
- The LLM agent dependencies are optional; the README separately demonstrates BrowserSession for direct navigation and screenshot capture.
- The CodeAgent runs generated Python in a persistent namespace, while browser control uses direct CDP communication.
- The repository documents MCP server and CLI daemon modes in addition to Python use.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Install OpenBrowser and prepare an LLM-backed agent
Set up the Python package and provider credentials needed for the documented CodeAgent example.
Before you start
- Python 3.12 or later.
- A Chromium-based browser such as Chrome, Edge, Brave or Chromium; the installer can use an installed browser, or install Chromium on first run when uvx is available.
- An API key for a supported LLM provider, such as Google, OpenAI or Anthropic.
Steps
Install the package with the command below.
pip install openbrowser-ai[agent]Set the provider's documented API-key environment variable, such as
GOOGLE_API_KEY.Use the README's Python example to create a
CodeAgentwith a task, such as searching for Python tutorials, and a provider model.Run the example with
asyncio.run(main())and print the returned result.
Expected outcome
The documented example runs a CodeAgent task and prints its result.
Guide limitations
- The guide follows the README example and does not establish successful execution or compatibility in a particular environment.
- A working provider credential and browser setup are needed for an agent workflow.
Use BrowserSession to navigate to a page and capture a screenshot
Set up a direct browser session that visits a URL and takes a screenshot.
Before you start
- Python 3.12 or later and the
openbrowser-aipackage. - An installed or automatically installed Chromium-based browser.
Steps
Install the package with the command below.
pip install openbrowser-aiCreate a
BrowserProfileand aBrowserSessionusing the documented Python example.Start the session, navigate to a URL, call
screenshot(), and stop the session.
Expected outcome
The example captures a screenshot after navigating to the specified page.
Guide limitations
- The example documents navigation and screenshot capture, not an LLM agent task.
- The guide does not verify successful execution in a particular environment.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 8 October 2026
- Original website title
- billy-enrizky/openbrowser-ai · Official homepage title · retained in its source language