MCP server Official
Firecrawl MCP Server
Firecrawl's MCP server to search, scrape, crawl and interact with the live web, hosted or self-run.
- Author
- Firecrawl
- Works with
- Any MCP client
- License
- MIT
- Documented
- Oct 2, 2026
What Firecrawl MCP Server is
An MCP server bringing Firecrawl to MCP-compatible agents for web search, scraping, crawling, page interaction and research. A keyless hosted tier exposes scrape, search and parse; more tools need an API key.
What it does
- Scrape a URL to markdown or schema JSON
- Search the web with optional content
- Crawl and map sites
- Interact with pages (click, type, navigate)
- Monitor pages for changes and check credit usage
When to use it
Good fit
- Giving an agent clean web content
- Known-URL scraping or site crawling
Not for
- Driving a browser step by step from the client: firecrawl_interact runs one turn per call
- Keyless hosted use of crawl, map and agent: need an API key
How to install Firecrawl MCP Server
You need: Node.js for npx; Firecrawl API key for full tool set
- Keyless: connect to https://mcp.firecrawl.dev/v2/mcp
- Or OAuth: https://mcp.firecrawl.dev/v2/mcp-oauth
- Or local: env FIRECRAWL_API_KEY=fc-YOUR_API_KEY npx -y firecrawl-mcp
- Add to your MCP client config
{
"mcpServers": {
"firecrawl-mcp": {
"command": "npx",
"args": [
"-y",
"firecrawl-mcp"
],
"env": {
"FIRECRAWL_API_KEY": "YOUR_API_KEY"
}
}
}
}Permissions and security
What it can access
- Sends URLs and queries to Firecrawl; fetches web content
Security notes from the maintainers
- API key set in env or Authorization header
Review the source before granting an agent access to credentials or production systems. We document what the maintainers state; we do not audit code.
Limitations
- Up to 26 tools in full profile; keyless endpoint exposes 3
Frequently asked questions
Related guides
Sources and gaps
- https://raw.githubusercontent.com/mendableai/firecrawl-mcp-server/main/README.md readme, checked Oct 2, 2026
- https://www.firecrawl.dev docs, checked Oct 2, 2026
Not stated by the maintainers: README fetched at mendableai path redirects to firecrawl org; last_release date not visible in the checked README.