Developer Tools

Firecrawl MCP Server

Connect your account, then chat with AI to run tools.

Firecrawl is a web scraping and crawling API that converts websites into clean, LLM-ready markdown. Provides single-page scraping, multi-page crawling, sitemap discovery, and structured data extraction.

Official server
Agent guide included
Start Chatting

Opens MCPBundles Studio with this server selected. After sign-in, chat and run tools from the same thread.

AI Skill
SKILL.md

Domain knowledge for Firecrawl — workflow patterns, data models, and gotchas for your AI agent.

Firecrawl transforms web content into clean, structured data optimized for AI consumption. It handles the complexity of modern web scraping including JavaScript rendering, anti-bot measures, and content cleaning to deliver consistent results.

Core Capabilities

Single-page extraction — Scrape individual URLs with advanced options for content filtering, format control, and metadata extraction. Returns clean markdown with optional structured data extraction using LLM processing.

Multi-page crawling — Launch crawl jobs that systematically extract content from entire websites or sections. Jobs run asynchronously with status checking and result retrieval.

Site discovery — Map websites to discover all indexed URLs, helping you understand site structure before crawling or identify specific pages to target.

Web search integration — Search the web and optionally extract content from results in a single operation, combining discovery and extraction workflows.

Interactive browsing — Scrape pages and then interact with them in live browser sessions (click elements, fill forms, navigate). Sessions must be explicitly stopped to free resources.

Autonomous research — Deploy AI agents that independently browse, search, and extract information based on research objectives. Separate agent jobs run asynchronously.

Domain Model

Content extraction supports multiple output formats (markdown, structured data, raw HTML) with configurable options for headers, links, media handling, and content filtering. Crawl and agent jobs follow async patterns requiring status polling until completion.

Key Considerations

Interactive browsing sessions consume server resources — always stop sessions when finished. Some tools are deprecated in favor of the scrape + interact workflow. Large crawl jobs may take time to complete, so implement appropriate polling intervals when checking status.

Frequently Asked Questions

What is the Firecrawl MCP server?

Firecrawl is a web scraping and crawling API that converts websites into clean, LLM-ready markdown. Provides single-page scraping, multi-page crawling, sitemap discovery, and structured data extraction. It provides tools that AI agents can use through the Model Context Protocol (MCP).

How do I connect Firecrawl to my AI agent?

Add the MCPBundles server URL to your MCP client configuration (Claude Desktop, Cursor, VS Code, etc.). The URL format is: https://mcp.mcpbundles.com/bundle/firecrawl-mcp. Authentication is handled automatically.

What authentication does Firecrawl require?

Firecrawl uses API Key. Firecrawl requires credentials. Connect via MCPBundles and authentication is handled automatically.

Setup Instructions

Connect Firecrawl to any MCP client in minutes

https://mcp.mcpbundles.com/bundle/firecrawl-mcp

What is MCP?

Model Context Protocol lets AI tools call external capabilities securely through a single URL. This bundle groups tools behind an MCP endpoint that many clients can use.

Use this bundle in 3 steps

  1. Copy the MCP URL above
  2. Open your AI tool and add a new MCP/connector
  3. Paste the URL and follow any auth prompts

Claude Desktop Users

Skip the manual setup! Use the .mcpb file format for one-click installation. Check the Claude Desktop tab for setup instructions.

Pick your tool tab for exact steps

Select ChatGPT, Cursor, Claude Code, or another tab for copy-paste config.

Ready to use Firecrawl?

Sign in to connect your credentials and start running tools from the chat.

Firecrawl MCP Server & Skill