Skip to content

Summary

Official Browserless MCP server: scrape, search, map and crawl sites, run Lighthouse audits and custom Puppeteer code, and drive persistent cloud browser sessions from any MCP client.

Features

  • Smart scraper returns pages as markdown, HTML or AI-extracted JSON, handling anti-bot measures
  • Web, news and image search with optional scraping of each result
  • Site map discovery and multi-page crawls with depth and path filters
  • Lighthouse performance audits and page export to PDF or images
  • Custom Puppeteer functions and a persistent browser agent with multi-tab support
  • Usage, saved login profiles, session replays and request logs from the same token

Installation

Set up this MCP server in your favorite AI agent — copy a ready-made configuration below.

Any MCP-compatible agent

Most agents (Claude, Cursor, Windsurf, VS Code, and more) read a standard mcpServers configuration.

  1. Open your agent's MCP configuration file.
  2. Merge the snippet below into it, filling in the headers with your own values.
  3. Restart the agent — the "Browserless MCP" tools will be available.
{
  "mcpServers": {
    "browserless-mcp": {
      "type": "http",
      "url": "https://mcp.browserless.io/mcp",
      "headers": {
        "Authorization": "Bearer <your-browserless-token>"
      }
    }
  }
}

Description

Browserless MCP is the official Model Context Protocol server for Browserless.io, the hosted headless-browser service. It turns the Browserless APIs into tools an agent can call, so a model can read JavaScript-heavy pages, collect data across a site or operate a real browser without the user running Chrome locally.

The scraping tools cover the common jobs: browserless_smartscraper returns one page as markdown, HTML or AI-extracted JSON and handles anti-bot measures; browserless_search runs web, news or image searches with geo and time filters and can scrape each result; browserless_map lists a site's URLs from sitemaps and links; and browserless_crawl scrapes every discovered page with depth and path filters. browserless_performance runs Lighthouse audits, browserless_export fetches a page as HTML, PDF or an image, and browserless_function executes custom Puppeteer JavaScript in the cloud. For interactive work, browserless_agent drives a persistent session in a snapshot-plan-act loop with multi-tab support, backed by on-demand recipes for cookie banners, modals, shadow DOM and captchas. Account, usage, saved login profiles, session replays and request logs are readable too.

The usual setup is the hosted streamable-HTTP endpoint at https://mcp.browserless.io/mcp, authenticated with a Browserless API token in an Authorization: Bearer header or a ?token= query parameter (for clients such as Claude.ai custom connectors that only take a URL). Browserless has a free tier. The server can also be self-hosted from the repository's Docker image and pointed at your own Browserless instance. The source is published under the SSPL-1.0 licence, not an OSI open-source licence.

New

Official Oxylabs MCP server: live web search, company lookup, and page scraping to Markdown or schema-defined JSON, including JavaScript-heavy and bot-protected sites, plus Oxylabs' dedicated scrapers.

New

Jina AI's official remote MCP server: read pages and PDFs as Markdown, search the web, arXiv and SSRN, take screenshots, rerank and deduplicate.

Official SerpApi MCP server: real-time Google, Bing, YouTube and other search results as structured JSON or token-saving markdown.

MCP: Crawlbase MCP

by Crawlbase

Official Crawlbase MCP server giving agents live web access — fetch any URL as raw HTML, clean Markdown or a screenshot, with JS rendering and proxy rotation handled for you.

Browse all MCP servers →