Deploy Steel Browser
Self-hosted browser API for AI agents, with an API key and persistent data
Steel Browser
Just deployed
/data
Deploy and Host Steel Browser on Railway
Steel Browser is an open-source browser API for AI agents and automation. It runs Chrome sessions you control over CDP with Puppeteer, Playwright or Selenium, offers quick actions such as scrape, screenshot and PDF, and includes a web UI to watch sessions live and read their logs.
About Hosting Steel Browser
This template runs Steel Browser v0.5.5-beta as one service built from a public GitHub repo, with a volume at /data for session logs, persisted browser profiles and session files. Steel Browser has no login of its own, so the template puts a small API-key gate in front of it: every request needs the generated STEEL_API_KEY, and only the health check is open. Railway gives the service an HTTPS domain, and session responses return matching wss:// URLs. In our tests it idled at about 0.4 GB of RAM and peaked near 0.9 GB; heavy pages need more. Each instance runs one browser at a time.
Common Use Cases
- Give AI agents a real browser to read and act on web pages
- Scrape JavaScript-heavy sites and turn pages into markdown, screenshots or PDFs
- Run Puppeteer, Playwright or Selenium scripts against a hosted Chrome instead of a local one
Dependencies for Steel Browser Hosting
- The official Steel Browser image (
ghcr.io/steel-dev/steel-browser, pinned by digest), with Chromium included - A Railway volume mounted at
/data
Deployment Dependencies
- Steel Browser source and docs: https://github.com/steel-dev/steel-browser
- Release v0.5.5-beta: https://github.com/steel-dev/steel-browser/releases/tag/v0.5.5-beta
- Template repo: https://github.com/aalfath/steel-browser-railway-template
Implementation Details
Find the key in the service's Variables tab (STEEL_API_KEY). Send it as the steel-api-key header, as a Bearer token, or as ?apiKey= on WebSocket URLs. In a browser, open https://YOUR-DOMAIN/ui and log in as user steel with the key as the password.
const s = await (await fetch("https://YOUR-DOMAIN/v1/sessions", {
method: "POST",
headers: { "steel-api-key": KEY, "content-type": "application/json" },
body: "{}",
})).json();
const ws = new URL("wss://YOUR-DOMAIN/");
ws.searchParams.set("apiKey", KEY);
ws.searchParams.set("sessionId", s.id);
const browser = await puppeteer.connect({ browserWSEndpoint: ws.href });
Optional settings such as PROXY_URL or DEFAULT_TIMEZONE can be added as service variables later. The repo README lists them.
Why Deploy Steel Browser on Railway?
Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.
By deploying Steel Browser on Railway, you are one step closer to supporting a complete full-stack application with minimal burden. Host your servers, databases, AI agents, and more on Railway.
Template Content
Steel Browser
aalfath/steel-browser-railway-template