Firecrawl
Free使用 Firecrawl 提取网页数据，支持 JavaScript 渲染。
FreeFree tier
Inputs: url
About Firecrawl
Firecrawl MCP Server is an open-source Model Context Protocol (MCP) server that enables AI agents (e.g., Claude, Cursor) to search, scrape, and interact with live websites. It supports JavaScript rendering, returns clean structured data, and includes features like autonomous deep research, automatic retries, and rate limiting. The server can be used via a hosted keyless free tier (rate-limited) or self-hosted. Tools like scrape, search, and interact work without an API key, while crawl, map, agent, and extract require one. It also offers SSE support and a separate read-only search-only endpoint.
Key Features
Search the web and get full page content
Scrape any URL into clean, structured data
Interact with pages — click, navigate, and operate
Deep research with autonomous agent
Automatic retries and rate limiting
Cloud and self-hosted support
SSE support
Keyless free tier for basic tools (scrape, search, interact)
Pros & Cons
Pros
- Open source and free to use (rate-limited free tier available)
- Supports JavaScript rendering for modern web pages
- Can be self-hosted for full control
- Works with popular AI clients like Claude and Cursor
- Includes deep research capabilities beyond basic scraping
Cons
- Advanced tools (crawl, map, agent, extract) require an API key
- Free tier is rate-limited
- Limited to MCP-compatible clients
- Setup requires some technical knowledge (e.g., npm, environment variables)
Best For
Providing real-time web data to AI chatbots and assistantsScraping and structuring website content for LLM contextAutomated web research for AI agentsIntegrating web interaction capabilities into MCP-compatible IDEs (e.g., Cursor, Claude)
FAQ
What is Firecrawl MCP Server?
It is an open-source Model Context Protocol (MCP) server that brings Firecrawl's web scraping and search capabilities to MCP-compatible AI agents like Claude and Cursor.
Is Firecrawl MCP Server free?
Yes, there is a keyless free tier that allows using scrape, search, and interact tools (rate-limited). Other tools require a Firecrawl API key.
What tools are available?
Available tools include firecrawl_scrape, firecrawl_search, firecrawl_interact, firecrawl_crawl, firecrawl_map, firecrawl_agent, firecrawl_extract, and deep research tools.
How do I run Firecrawl MCP Server?
You can use the hosted endpoint (keyless or with API key) or install via npm with npx firecrawl-mcp. Configuration instructions for Cursor are provided in the repository.
What is the search-only endpoint?
A separate hosted endpoint (https://mcp.firecrawl.dev/v2/mcp-search) that exposes only read-only search and research tools, requiring OAuth and performing no page-content fetching.