The Steel browser provides a REST API to control a headless browser powered by Puppeteer, managing browser instances, sessions, and pages. It offers three main ways for browser automation: Quick Actions API for simple, read-only jobs; Browser Sessions API for complex, stateful workflows; and Selenium Integration for teams with existing Selenium workflows. The article also provides detailed instructions on how to get started with Steel, including building and running the Docker image, and using the REST API. Contributions to the open-source project are welcomed, and the license is Apache 2.0.
Key takeaways:
- Steel is an open-source browser API designed for AI developers to programmatically control browsers that work with their favorite frameworks.
- Steel provides a REST API to control a headless browser powered by Puppeteer, managing browser instances, sessions, and pages, and allowing complex browsing tasks to be performed programmatically.
- Steel offers full browser control, session management, proxy support, extension support, debugging tools, anti-detection, and resource management.
- Users can get started with Steel by creating a Steel Cloud account, and the browser can be run locally with GHCR, or deployed to Railway or Render with a single click.