Steel Browser

Steel Browser is an open-source browser automation API and sandbox, hosted on GitHub (steel-dev/steel-browser), purpose-built for AI agents and applications that need to interact with the web. Instead of an agent developer having to manage Chrome processes, sessions, and infrastructure themselves, Steel provides a managed browser instance controlled over the Chrome DevTools Protocol, accessible through Puppeteer, Playwright, or Selenium. It handles persistent session state, including cookies and local storage, across requests, supports proxy chains for IP rotation, allows loading custom Chrome extensions, and includes stealth/anti-detection plugins and fingerprint management for scraping and automation use cases. It also ships utility endpoints for converting pages to markdown, taking screenshots, and generating PDFs, plus built-in request logging and a debugging UI for inspecting live sessions. The project is split into separate API and UI components and can be self-hosted via Docker, deployed to platforms like Railway or Render, or run locally with Node.js and Chrome. It's aimed at developers building AI agents, browser-automation tools, or scraping pipelines who want production-grade browser infrastructure without operating it themselves, and it has an active open-source community with roughly 7,400 GitHub stars.