Skip to main content

Browser

Abstra provides a built-in browser environment in the cloud, powered by Selenium Grid. You can use it for RPAs, web scraping, automated testing, and any task that requires a browser.

The browser environment supports both Selenium (WebDriver protocol) and Playwright (CDP protocol), so you can choose whichever library fits your use case.

How it works​

When your project is deployed, Abstra automatically provisions a Chrome browser instance for your project. It scales to zero when not in use and starts up on demand.

  • Selenium connects via the ABSTRA_SELENIUM_URL environment variable
  • Playwright connects automatically via the SELENIUM_REMOTE_URL environment variable (no code changes needed)

In local development, both libraries fall back to launching a local Chrome instance.

A few things to keep in mind:

  • The playwright package ships with the project runtime — don't add it to requirements.txt and don't run playwright install. Other browser libraries (e.g. selenium) must be added to requirements.txt as usual.
  • In deployed projects Chrome runs on a separate machine from your code: it cannot open local files (file:// URLs) — feed it URLs or content directly (e.g. page.goto, page.set_content). Files produced by the browser (PDFs, screenshots, downloads) are written locally where your code runs.
Using the browser with agents

You can give an AI agent a headless browser using BrowserTools. The agent can navigate, click, fill forms, take screenshots, and extract page data autonomously. See Agent Tools for details.