Browserless

aid: browserless
name: Browserless
description: >-
  Browserless is a cloud browser-automation platform that runs managed
  Chromium, Chrome, Edge, Firefox, and WebKit browsers for Puppeteer,
  Playwright, and Selenium clients, plus a GraphQL-based stealth automation
  layer (BrowserQL) and a family of REST APIs for screenshots, PDFs, content
  scraping, function execution, performance audits, smart scraping, search,
  mapping, and full-site crawling. A built-in MCP server exposes browser
  tooling to Claude, Cursor, VS Code, and other agentic clients. The company
  also maintains a popular open-source Docker image of the same name (13k+
  GitHub stars), residential proxies, CAPTCHA solving, session profiles,
  recordings, and hybrid live-URL workflows. Target customers are AI
  startups, agent developers, web scrapers, RPA teams, QA / monitoring teams,
  and enterprises running production browser automation at scale, with
  free, prototyping, starter, scale, and enterprise plans (including
  self-hosted licensing).
type: Index
position: Provider
access: 3rd-Party
image: https://kinlane-images.s3.amazonaws.com/shared/apis-json/apis-json-logo.jpg
tags:
- Headless Browser
- Browser Infrastructure
- Web Automation
- AI Agents
- Web Scraping
- BrowserQL
- Puppeteer
- Playwright
- Selenium
- CDP
- Stealth
- CAPTCHA Solving
- Residential Proxy
- PDF Generation
- Screenshots
- Smart Scrape
- Crawl
- Search
- MCP
- Session Recording
- Hybrid Automation
url: https://raw.githubusercontent.com/api-evangelist/browserless/refs/heads/main/apis.yml
created: '2026-05-25'
modified: '2026-05-25'
specificationVersion: '0.20'
apis:
- aid: browserless:browser-api
  name: Browserless Cloud API
  description: >-
    The Browserless Cloud API is a multi-surface browser platform combining
    (1) WebSocket BaaS endpoints compatible with Puppeteer, Playwright, and
    Selenium across Chromium, Chrome, Edge, Firefox, and WebKit; (2) a
    GraphQL automation layer called BrowserQL for stealth scraping with
    built-in CAPTCHA solving and residential proxies; and (3) a REST API
    family for screenshots, PDFs, content, scrape, function execution,
    performance audits, smart-scrape, search, map, crawl, unblock, and
    session management. All surfaces share a single API token and a
    regional production-{sfo,lon,ams} host pattern.
  humanURL: https://docs.browserless.io
  baseURL: https://production-sfo.browserless.io
  tags:
  - BaaS
  - BrowserQL
  - REST
  - Screenshots
  - PDF
  - Scrape
  - Smart Scrape
  - Search
  - Crawl
  - Map
  - Unblock
  - Function
  - Performance
  - Session
  - Profile
  - Proxy
  - CAPTCHA
  - Stealth
  - MCP
  properties:
  - type: Documentation
    url: https://docs.browserless.io
  - type: APIReference
    url: https://docs.browserless.io/open-api
  - type: GettingStarted
    url: https://docs.browserless.io/overview/quick-start
  - type: SignUp
    url: https://account.browserless.io/signup
  - type: SDK
    url: https://github.com/puppeteer/puppeteer
  - type: SDK
    url: https://github.com/microsoft/playwright
  - type: SDK
    url: https://github.com/SeleniumHQ/selenium
  - type: GitHubRepository
    url: https://github.com/browserless/browserless
  - type: GitHubRepository
    url: https://github.com/browserless/browserless-mcp
  - type: GitHubRepository
    url: https://github.com/browserless/claude-plugin
  - type: GitHubRepository
    url: https://github.com/browserless/gemini-extension
  - type: GitHubRepository
    url: https://github.com/browserless/n8n-nodes-browserless-api
  - type: GitHubRepository
    url: https://github.com/browserless/dify-plugin
  - type: GitHubRepository
    url: https://github.com/browserless/cursor-mcp-plugin
  - type: GitHubRepository
    url: https://github.com/browserless/browserless-langchain
  - type: GitHubRepository
    url: https://github.com/browserless/browserless-llamaindex
  - type: GitHubRepository
    url: https://github.com/browserless/browserless-agno
  - type: GitHubRepository
    url: https://github.com/browserless/bolt-diy-browserless-mcp-starter
  - type: OpenAPI
    url: openapi/browserless-openapi.yml
  - type: JSONSchema
    url: json-schema/browserless-session-schema.json
  - type: JSONStructure
    url: json-structure/browserless-session-structure.json
  - type: JSONLD
    url: json-ld/browserless-context.jsonld
  - type: Examples
    url: examples/browserless-screenshot-example.json
  - type: Plans
    url: plans/browserless-plans-pricing.yml
  - type: RateLimits
    url: rate-limits/browserless-rate-limits.yml
  - type: FinOps
    url: finops/browserless-finops.yml
  - type: Capabilities
    url: capabilities/rest-apis.yaml
  - type: Capabilities
    url: capabilities/browserql.yaml
  - type: Capabilities
    url: capabilities/sessions.yaml
  - type: Capabilities
    url: capabilities/profiles.yaml
  - type: Vocabulary
    url: vocabulary/browserless-vocabulary.yml
  - type: Rules
    url: rules/browserless-rules.yml
  - type: ChangeLog
    url: https://www.browserless.io/changelog
  - type: StatusPage
    url: https://status.browserless.io
  - type: Pricing
    url: https://www.browserless.io/pricing
  - type: Enterprise
    url: https://www.browserless.io/enterprise
  - type: MCP
    url: https://mcp.browserless.io/mcp
  - url: graphql/browserless-graphql.md
    type: GraphQL
  features:
  - name: Browsers-as-a-Service (BaaS)
    description: >-
      WebSocket endpoints (production-sfo, production-lon, production-ams)
      that drop into existing Puppeteer (connect) and Playwright
      (connectOverCDP / connect) code with one URL change. Selenium-grid
      compatible endpoints exist alongside.
  - name: BrowserQL (BQL)
    description: >-
      A GraphQL automation language with goto, click, type, evaluate,
      waitForSelector, html, mapSelector, screenshot, pdf, and reconnect
      mutations. Stealth (/stealth/bql) and Chrome (/chrome/bql) routes
      provide hardened fingerprints and built-in CAPTCHA solving.
  - name: REST API Family
    description: >-
      /screenshot, /pdf, /content, /scrape, /function, /performance,
      /download, /export, /search, /smart-scrape, /map, /crawl, and
      /unblock endpoints over chrome, chromium, and edge prefixes for
      one-shot HTTP usage without a browser library.
  - name: Stealth and Bot Detection
    description: >-
      Path-based stealth routes, WebGL fingerprinting, UA spoofing, and
      entropy injection. Supports reCAPTCHA, hCaptcha, DataDome, Lemin,
      MTCaptcha, and slider challenges.
  - name: Residential Proxies
    description: >-
      Built-in residential proxy routing with country / city / sticky
      session controls, plus support for external bring-your-own proxies.
  - name: Session Profiles
    description: >-
      Reusable, encrypted browser profiles (/profile, /profiles) that
      persist cookies, localStorage, and authenticated state across runs,
      including authenticated-profile reuse via query parameters.
  - name: Hybrid Automation / LiveURL
    description: >-
      Browserless.liveURL CDP commands let a script hand control of a
      running session to a human (or a watching agent) over an embedded
      live URL, then reclaim control.
  - name: Session Recording and Replay
    description: >-
      rrweb-based session replay plus WebM screen recording for debugging
      and audit, with replay storage from 1 to 90 days by plan.
  - name: Smart Scrape, Search, Map, Crawl
    description: >-
      AI-tuned data-extraction REST endpoints that pick the cheapest
      strategy (HTTP fetch, proxy, headless, CAPTCHA), perform web
      search, build site maps, and run depth-bounded full-site crawls.
  - name: Official MCP Server
    description: >-
      mcp.browserless.io/mcp exposes browser-agent, smart-scraper, custom
      code, downloads, page-export, Lighthouse, and crawl tools to Claude
      Desktop, Cursor, VS Code, Windsurf, and any MCP client.
  - name: Open Source Core
    description: >-
      The browserless/browserless Docker image (Apache 2.0, 13k+ stars)
      is free for non-commercial use and underpins the cloud service,
      enabling self-hosting and private deployments.
  - name: Enterprise Self-Hosting
    description: >-
      Enterprise tier supports private cloud deployments, licensed
      self-hosting, GPU infrastructure, SSO, dedicated account
      management, and millions of units of monthly capacity.
  useCases:
  - name: AI Agent Browsing
    description: >-
      Give Claude, GPT, or open-source agents a real browser to research,
      fill forms, click flows, solve CAPTCHAs, and return structured
      data.
  - name: Stealth Web Scraping at Scale
    description: >-
      Run fleets of fingerprint-randomized Chromium browsers behind
      residential proxies for production data extraction.
  - name: PDF and Screenshot Generation
    description: >-
      High-fidelity PDF and image rendering from HTML or URL inputs, with
      control over viewport, device emulation, headers, and cookies.
  - name: End-to-End Testing
    description: >-
      Drive Playwright, Puppeteer, and Selenium suites against staging
      and production in managed browsers without maintaining grid
      infrastructure.
  - name: RPA and Hybrid Human-in-the-Loop Workflows
    description: >-
      Combine programmatic automation with embedded LiveURL handoff so a
      human (or supervising agent) can complete steps a script cannot.
  - name: Synthetic Monitoring and Lighthouse Audits
    description: >-
      Schedule performance, accessibility, SEO, and best-practices audits
      across a fleet of pages and capture visual / network artifacts.
  - name: AI-Tuned Site Crawls and Knowledge Building
    description: >-
      Use /crawl, /map, and /search to feed agents with structured site
      context for RAG pipelines and research agents.
  integrations:
  - name: Puppeteer
  - name: Playwright
  - name: Selenium
  - name: Browser Use
  - name: Stagehand
  - name: LangChain
  - name: LlamaIndex
  - name: Agno
  - name: Mastra
  - name: Vercel AI SDK
  - name: n8n
  - name: Make
  - name: Zapier
  - name: Dify
  - name: bolt.diy
  - name: Cursor
  - name: VS Code
  - name: Windsurf
  - name: Claude Desktop
  - name: Gemini CLI
  - name: Anthropic
    url: https://github.com/api-evangelist/anthropic
  - name: MCP
  authentication:
  - type: API Key
    description: >-
      A single API token authenticates all surfaces. REST and BaaS pass
      it as a `?token=YOUR_API_TOKEN` query parameter on the connection
      URL; BrowserQL accepts the same token in the URL or via header.
common:
- type: PostmanWorkspace
  url: https://www.postman.com/kinlaneapi/browserless/overview
- type: ArazzoWorkflows
  url: arazzo/
  workflows:
  - url: arazzo/browserless-content-to-pdf-workflow.yml
    name: Browserless Content To PDF
    summary: Render a URL's HTML, then feed that exact HTML into the PDF engine to produce a self-contained PDF.
  - url: arazzo/browserless-full-page-archive-workflow.yml
    name: Browserless Full Page Archive
    summary: Unblock a protected URL, then branch into structured scraping or a content+PDF archive depending on whether unblocked HTML was returned.
  - url: arazzo/browserless-function-download-workflow.yml
    name: Browserless Function Then Download
    summary: Use the function API to drive a page that triggers a file download, then retrieve the downloaded artifacts via the download API.
  - url: arazzo/browserless-function-then-screenshot-workflow.yml
    name: Browserless Function Then Screenshot
    summary: Run custom JavaScript against a page via the function API, then capture a screenshot to confirm the resulting state.
  - url: arazzo/browserless-performance-audit-workflow.yml
    name: Browserless Performance Audit
    summary: Run a Lighthouse-style performance audit on a URL, then capture a screenshot of the audited page.
  - url: arazzo/browserless-scrape-then-screenshot-workflow.yml
    name: Browserless Scrape Then Screenshot
    summary: Extract structured elements from a page, then capture a screenshot of the same page as visual evidence.
  - url: arazzo/browserless-site-capture-bundle-workflow.yml
    name: Browserless Site Capture Bundle
    summary: "Render a single URL three ways — HTML content, PNG screenshot, and PDF — in one pass."
  - url: arazzo/browserless-unblock-then-render-workflow.yml
    name: Browserless Unblock Then Render
    summary: Bypass bot detection on a protected URL, then re-render its HTML content and a PDF using the unblocked content.
- type: Website
  url: https://www.browserless.io
- type: Documentation
  url: https://docs.browserless.io
- type: APIReference
  url: https://docs.browserless.io/open-api
- type: Blog
  url: https://www.browserless.io/blog
- type: GitHubOrganization
  url: https://github.com/browserless
- type: Pricing
  url: https://www.browserless.io/pricing
- type: Enterprise
  url: https://www.browserless.io/enterprise
- type: ChangeLog
  url: https://www.browserless.io/changelog
- type: StatusPage
  url: https://status.browserless.io
- type: SignUp
  url: https://account.browserless.io/signup
- type: SignIn
  url: https://account.browserless.io
- type: TermsOfService
  url: https://www.browserless.io/terms
- type: PrivacyPolicy
  url: https://www.browserless.io/privacy
- type: LLMsTxt
  url: https://docs.browserless.io/llms.txt
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com