Assembly Line

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.

Steel Browser is an open-source browser automation project on GitHub under the Apache-2.0 license (permissive), with 7.9k stars. It is built with TypeScript and Fastify. Its last push was yesterday.

Stars
7.9k
License
Apache-2.0
Activity
Active
Latest release
v0.5.5-beta

Code on GitHub Website

Judgement

What it is
Building block
What it does
A self-hostable browser API that runs Chrome sessions with stored cookies, proxies and extensions, which you connect to with Puppeteer, Playwright or Selenium, plus a session viewer.
Adopt it when
You want to run your own pool of remote browser sessions for agents or scrapers instead of paying for a hosted browser service.
Watch out
Releases are still tagged beta, and the company's hosted Steel cloud gets features first, so the open-source server can lag.
Ready for
Usable
Runs as
Self-hosted server and cloud service
Built with
TypeScript, Fastify, Puppeteer and Chrome
How established
Established: started 2 years ago, latest release v0.5.5-beta.
What the license allows
You can build it into a product you sell, as long as you keep its copyright notice.

Judged on 2026-11-06 from the project's README and GitHub data. The license and age lines are worked out directly.

Built from

Follow it down: Steel Browser → TypeScript → Go → runs on a microchip

How compilers were built, back to punched paper tape

Open-source alternatives to Steel Browser

  1. 1
    Browser Use browser-use

    Open-source Python library for AI browser agents, plus Browser Use Cloud: stealth browsers and a hosted agent API.

    Building block Usable You are building an AI agent that has to complete tasks on websites that have no API, and you accept slower, less predictable runs than a scripted Playwright test.

    ★ 120k MIT Python Active, pushed today
  2. 2
    Playwright microsoft

    Playwright is a framework for Web Testing and Automation. It allows testing Chromium, Firefox and WebKit with a single API.

    Developer tool Production-ready You need reliable end-to-end tests across all major browser engines, or browser automation for scripts and AI agents.

    ★ 99k Apache-2.0 TypeScript Active, pushed today
  3. 3
    Puppeteer puppeteer

    JavaScript API for Chrome and Firefox

    Developer tool Production-ready You automate Chrome from JavaScript for scraping, PDFs or screenshots and do not need Safari's engine.

    ★ 96k Apache-2.0 TypeScript Active, pushed today
  4. 4
    Cypress cypress-io

    Fast, easy and reliable testing for anything that runs in a browser.

    Developer tool Production-ready Your front-end team wants an easy-to-debug browser test tool with an interactive runner, mainly for Chromium-family browsers.

    ★ 51k MIT TypeScript Active, pushed today
  5. 5
    Lightpanda lightpanda-io

    Lightpanda: the headless browser designed for AI and automation

    Developer tool Usable You run many headless sessions for scraping or AI agents and want far less memory and startup time than headless Chrome.

    ★ 41k AGPL-3.0 Zig Active, pushed today
  6. 6
    Selenium SeleniumHQ

    A browser automation framework and ecosystem.

    Developer tool Production-ready You need to drive real browsers through the WebDriver standard, use languages beyond Playwright's, or run a large existing Selenium suite or Grid.

    ★ 35k Apache-2.0 Java Active, pushed today
  7. 7
    CloakBrowser CloakHQ

    Stealth Chromium that passes every bot detection test. Drop-in Playwright replacement with source-level fingerprint patches. 50+ tests passed.

    Developer tool Usable Your scraping or agent runs are blocked by bot detection that catches stock headless Chrome and you have a legitimate right to the data.

    ★ 33k MIT Python Active, pushed 2 days ago
  8. 8
    ArchiveBox ArchiveBox

    🗃 Open source self-hosted web archiving. Takes URLs/browser history/bookmarks/Pocket/Pinboard/etc., saves HTML, JS, PDFs, media, and more...

    Complete system Usable You want your own lasting copies of web pages, for research, legal evidence or bookmarks, rather than relying on the Wayback Machine.

    ★ 29k MIT Python Active, pushed 2 days ago

Stars, license and activity come from GitHub and refresh on a rolling hourly schedule. The project's own description is quoted as written.