# How to connect the Firecrawl MCP server

To connect the official Firecrawl MCP server, add https://mcp.firecrawl.dev/v2/mcp to your MCP client as a Streamable HTTP server, then send a bearer token in the Authorization header. Firecrawl documents setup for VS Code.

- Endpoint: https://mcp.firecrawl.dev/v2/mcp
- Transport: Streamable HTTP
- Registry: io.github.firecrawl/firecrawl-mcp-server 3.37.0
- Publisher documentation reviewed: 2026-11-05 (connection tested)
- Guide page: https://apitracker.io/mcp-guides/firecrawl-mcp

## Before you start

- A Firecrawl account and API key for the full tool set and higher limits. Without a key the endpoint offers only search, scrape and parse, rate limited. Which tools a key unlocks depends on your plan, deployment and team policy.

## Steps

1. **Open Firecrawl's MCP server documentation** <https://docs.firecrawl.dev/mcp-server>
2. **Add the Firecrawl MCP server to your client** Add https://mcp.firecrawl.dev/v2/mcp as a remote MCP server that uses the Streamable HTTP transport (often labelled "http"), and send your Firecrawl API key as a bearer token in the Authorization header. Keep the key in an environment variable or secret store, not in shared configuration.
3. **Authorization (your Firecrawl API key)**
4. **To sign in instead of using a key, use Firecrawl's OAuth endpoint** Add https://mcp.firecrawl.dev/v2/mcp-oauth with no headers. Your client opens a browser where you sign in to Firecrawl and approve a team. Firecrawl says signing in and adding an API key give the same access. You can review and revoke connections in Firecrawl's MCP settings. Firecrawl's Claude, VS Code and Copilot CLI guides use this endpoint by default.
5. **To try it without an account, connect without a key** The same https://mcp.firecrawl.dev/v2/mcp endpoint works without credentials. Keyless connections get search, scrape and parse only, with rate limits.
6. **List the server's tools**

## Questions

### What is the Firecrawl MCP server URL?

The official Firecrawl MCP server is at https://mcp.firecrawl.dev/v2/mcp and uses the Streamable HTTP transport.

### How do I authenticate with the Firecrawl MCP server?

Send a bearer token in the Authorization header.

### Which AI clients can use the Firecrawl MCP server?

Firecrawl's documentation covers VS Code. Other clients that support remote Streamable HTTP MCP servers may also work.

### What do I need before connecting Firecrawl MCP?

A Firecrawl account and API key for the full tool set and higher limits. Without a key the endpoint offers only search, scrape and parse, rate limited. Which tools a key unlocks depends on your plan, deployment and team policy.

### Has this Firecrawl MCP guide been tested?

Firecrawl's documentation was reviewed on 2026-11-05. The connection was tested.

## Sources

- https://docs.firecrawl.dev/mcp-server (checked 2026-11-05)
- https://docs.firecrawl.dev/mcp-server/keyless (checked 2026-11-05)
- https://docs.firecrawl.dev/quickstarts/github-copilot (checked 2026-11-05)
- https://docs.firecrawl.dev/quickstarts/claude-desktop (checked 2026-11-05)
- https://registry.modelcontextprotocol.io/v0.1/servers/io.github.firecrawl%2Ffirecrawl-mcp-server/versions/3.37.0 (checked 2026-11-05)

Guide JSON: https://github.com/apideck-libraries/mcp-connect/blob/626f6a146170ded34891b30dca3af26f09e07466/proposals/connection-guide/0.1/reviewed/guides/firecrawl-mcp.json