steel-browser

Overview of steel-browser

What it is

Steel is an open-source browser API designed for AI agents and applications. It provides a batteries-included browser sandbox that handles session management, page control, and browser processes, allowing developers to automate web interactions without managing infrastructure. Under the hood, it uses Puppeteer and the Chrome DevTools Protocol (CDP) to control Chrome instances, and it is compatible with Puppeteer, Playwright, and Selenium.

Key features

• Full browser control via Puppeteer and CDP, with support for Puppeteer, Playwright, and Selenium
• Session management that maintains browser state, cookies, and local storage across requests
• Built-in proxy chain management for IP rotation
• Custom Chrome extension support
• Debugging tools including request logging and a UI for viewing/debugging sessions
• Anti-detection features through stealth plugins and fingerprint management
• Automatic cleanup and browser lifecycle management
• APIs to convert pages to markdown, readability, screenshots, or PDFs

Use cases

Steel is suitable for building live web agents, browser automation tools, and AI applications that need to interact with web pages programmatically. It can be used for scraping, taking screenshots, generating PDFs, or performing complex, stateful browsing tasks.

Getting started

The easiest way to start is by creating a Steel Cloud account. To run locally using Docker:bashdocker run -p 3000:3000 -p 9223:9223 ghcr.io/steel-dev/steel-browserAlternatively, run the API and UI separately:bashdocker compose upFor Mac Silicon, use:bashDOCKER_DEFAULT_PLATFORM=linux/arm64 docker compose upIf you have Node.js and Chrome installed, you can also run directly:bashnpm installnpm run devThis starts the Steel server on port 3000 and the UI on port 5173. For more details, see the repository README.

License and pricing

Licensed under Apache-2.0. The project is open source and free to use.

Based on the project's public README and repository metadata.