# 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.

Repository: https://github.com/steel-dev/steel-browser
Canonical: https://ross.abutalabs.com/products/steel-browser
Homepage: https://steel.dev
Language: TypeScript
License: Apache-2.0
License Family: permissive
Topics: ai, ai-agents, ai-tools, browser-automation, llm
Last push: 2026-08-25T19:41:14+00:00
Link (homepage): https://steel.dev
Link (site_page): https://docs.steel.dev/

## Health v2 (maintenance only)
Score: 89/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 99, longevity 47
- inputs: {"age_days": 670, "days_push": 8, "days_rel": 8, "gap_med": 10.5, "n_releases_24m": 23}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 7545, forks 972 (observed 2026-08-28T04:10:01.448877+00:00)

## What it is
Steel Browser is an open-source browser API and sandbox that manages browser sessions, proxies, stealth, and lifecycle so developers can build AI agents and web automation without infrastructure work. It exposes a REST API and SDKs (TypeScript, Python, Rust, Go) and connects with Puppeteer, Playwright, Selenium, and computer-use agents.

## Use cases
- run cloud browser sessions for AI agents
- scrape websites with stealth and proxy rotation
- convert web pages to markdown, screenshots, or PDFs
- connect Claude or Gemini computer use to a real browser
- maintain cookies and login state across automation runs
- solve CAPTCHAs during automated browsing
- self-host a browser automation API instead of building one

## When to choose
- you need managed, scalable browser infrastructure for LLM agents or web automation
- you want anti-detection, proxies, and session persistence out of the box
- you want to self-host a browser API compatible with Puppeteer/Playwright/Selenium

## When to avoid
- you only need simple local scraping with a single script
- you need a lightweight headless browser library rather than a full service
- you require a fully stable GA product - it is still in public beta

## Facets
- artifact type: service
- maturity: active
- function: web-scraping, http-server, browser-extension, workflow-automation, agent-framework
- domain: web-development, developer-tools, crawlers
- platform: self-hosted, cross-platform
- tags: browser-api, browser-automation, ai-agents, puppeteer, playwright, stealth, session-management, captcha-solving, proxy-support, computer-use, automation, nodejs, docker, web-server

## Member repositories
- steel-dev/steel-browser (main) score 89
- steel-dev/awesome-web-agents (docs) score 64

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:01.448877+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:37:24.568253+00:00, confidence not recorded.
  - readme: https://github.com/steel-dev/steel-browser (fetched 2026-08-28T04:10:01.448877+00:00, sha a36abd293ba7)
  - homepage: https://steel.dev (fetched 2026-08-29T08:32:47.864927+00:00, sha 4ff1b40cc4ed)
  - site_page: https://docs.steel.dev/ (fetched 2026-08-29T08:32:47.874539+00:00, sha c2b607837325)
- Data as of 2026-08-30T08:39:29.467469+00:00.