steel-dev/steel-browser

🔥 Open Source Browser API for AI Agents & Apps. Steel Browser is a batteries-included browser sandbox that lets you automate the web without worrying about infrastructure.

What it solves

Steel simplifies the development of AI agents and web automation tools by providing a managed browser API. It removes the need to build complex browser infrastructure from scratch, handling the heavy lifting of managing browser processes, sessions, and state so you can focus on building your application.

How it works

Steel acts as a server that manages Chrome instances. You can interact with it via a REST API, or by using the provided Python and Node.js SDKs. It supports three main interaction modes: persistent sessions for complex workflows, quick-action endpoints for tasks like scraping or taking screenshots, and an experimental commands API designed specifically for AI agents to navigate and interact with web pages.

Who it’s for

It is designed for developers and AI engineers who need to build reliable, scalable web agents or automation tools that require browser-based interaction, such as scraping, form filling, or complex web navigation.

Highlights

  • Full browser control via Puppeteer, Playwright, or Selenium.
  • Built-in session management for cookies, local storage, and state.
  • Integrated proxy management for IP rotation and anti-detection features.
  • Support for loading custom Chrome extensions.
  • Built-in debugging tools, including request logging and a session UI.
  • Ready-to-use endpoints for converting web pages to markdown, PDFs, or screenshots.

Related

  • Project
    SawyerHood/dev-browser

    A browser automation tool that allows AI coding agents to maintain persistent browser sessions and use compact accessibility trees for efficient web interaction.

  • Project
    Bin-Huang/camoufox-cli

    A command-line interface and set of skills for AI agents to browse the web using an anti-detect browser that spoofs fingerprints to avoid bot detection.

  • Project
    eze-is/web-access

    A browser automation skill for AI agents that provides web connectivity, CDP-based browser control, and the ability to use existing browser sessions and local history.

  • Project
    goodnight000/api-anything

    A tool that teaches AI agents to interact with websites via HTTP requests instead of browser automation, reducing latency and token costs.

  • Project
    sleepinginsummer/agent-browser-cli

    A CLI tool that connects AI Agents to a real Chrome browser session, allowing them to reuse login states and perform precise page control and perception.