Skip to content

Browserless

Category: Infra (Cloud Browser)

Official site: browserless.io

GitHub: github.com/browserless/browserless

Docs: docs.browserless.io

What it is

Browserless is a cloud browser automation platform that lets you connect Puppeteer, Playwright, or Selenium to managed headless Chromium, Firefox, or WebKit browsers over WebSocket. Instead of launching a local browser process, you point your existing automation script at a remote WebSocket endpoint. Browserless handles the browser lifecycle, concurrency, stealth routing, CAPTCHA solving, and proxy access on managed infrastructure.

It also ships a BrowserQL interface (GraphQL-based) for structured scraping with conditional logic, REST APIs for one-shot operations (screenshots, PDFs, content extraction), and an MCP server endpoint for AI agent integrations.

Free vs paid

TierPriceConcurrencyUnits/monthMax session
Free$0/mo (no card)2 browsers1,0001 minute
Prototyping$25/mohighermoreup to 10 min
Scalecustomhighermoreup to 60 min

A unit is up to 30 seconds of browser connection time. Additional unit costs apply for proxy usage (6 units/MB residential, 2 units/MB datacenter) and CAPTCHA solving (10 units per successful solve).

Regional endpoints

RegionWebSocket URL
US West (San Francisco)wss://production-sfo.browserless.io
Europe (London)wss://production-lon.browserless.io
Europe (Amsterdam)wss://production-ams.browserless.io

Key capabilities

  • Drop-in replacement for local browser launch in Puppeteer and Playwright scripts
  • Stealth mode via /chromium/stealth path (not automatic)
  • CAPTCHA auto-solve via solveCaptchas=true connection param (costs 10 units per solve)
  • Persistent sessions with configurable storage retention (1 day on free tier)
  • MCP server at https://mcp.browserless.io/mcp for Claude and LangChain integrations
  • Self-hostable via Docker

Safety and ToS note

Browserless is a legitimate managed infrastructure service. Stealth and CAPTCHA-solving features are intended for authorized automation. Users are responsible for complying with the terms of service of any site they automate against. The Browserless ToS prohibits using the service for unauthorized scraping or access. CAPTCHA solving costs units and is not automatic; you must pass solveCaptchas=true explicitly.