Firecrawl MCP Server

A Model Context Protocol (MCP) server implementation that integrates Firecrawl for searching, scraping, and interacting with the web. Our MCP server is open-source and available on GitHub.

Features

Search the web and get full page content
Scrape any URL into clean, structured data
Parse local files such as PDFs, DOCX, XLSX, and HTML
Interact with pages — click, navigate, and operate
Deep research with autonomous agent
Cloud and self-hosted support
Streamable HTTP support

Choose how you want to start

The dedicated account endpoint and recovery behavior described below are not live yet. Today, use /v2/mcp without credentials or with an API key in the Authorization header.

Try it now

Connect to the hosted keyless endpoint. No account or API key is required. Keyless mode exposes exactly Search, Scrape, and Parse and is rate-limited per IP.

Connect your account

With the account-endpoint rollout, interactive connectors use a dedicated URL. Your client opens a browser so you can sign in, choose a team, approve access, and revoke the connection later.

Run unattended

Use an API key for CI, servers, and other workloads that run without you. Send the key in the Authorization header, never in the URL.

Try it now

Local coding agents such as Claude Code, Codex, Cursor, and OpenCode should start here:

https://mcp.firecrawl.dev/v2/mcp

Keyless connections expose Search, Scrape, and Parse. If a request needs an account-only tool, Firecrawl explains how to connect an account or add an API key. Quota errors include retry guidance. If a configured credential is invalid, Firecrawl returns a credential error instead of silently falling back to keyless mode. Expired or revoked OAuth connections ask compatible clients to reconnect.

Connect your account

After the account-endpoint rollout is live, ChatGPT, Claude web/desktop connectors, and other account-attached integrations should use:

https://mcp.firecrawl.dev/v2/mcp-oauth

This endpoint will have its own OAuth resource identity and will always ask an anonymous client to sign in. Compatible clients identify themselves automatically using Client ID Metadata Documents or Dynamic Client Registration, so users do not need to create an OAuth client manually. The endpoint also accepts Firecrawl API keys in an Authorization header.

Run unattended

Send a Firecrawl API key to either hosted endpoint as a Bearer header. Keep the key in an environment variable and use the client’s native environment binding rather than writing it into a shared configuration file. For example, the Firecrawl CLI can configure Codex without persisting the key:

export FIRECRAWL_API_KEY=fc-your-api-key
firecrawl setup mcp --agent codex --global --yes

Compatible existing OAuth tokens explicitly bound to https://mcp.firecrawl.dev/v2/mcp remain accepted there. Tokens with a missing or ambiguous audience fail closed. After rollout, new interactive connections should use the account endpoint above. Audience compatibility is intentionally one-way during rollout: compatible existing tokens explicitly bound to /v2/mcp also work on /v2/mcp-oauth, while new /v2/mcp-oauth tokens do not work on /v2/mcp. This preserves compatible existing connections without weakening the resource binding of newly issued account tokens. Existing https://mcp.firecrawl.dev/<API_KEY>/v2/mcp configurations also continue working during the migration. New configurations should use OAuth or a Bearer header so the credential does not appear in the URL.

Upgrade from keyless mode

When a keyless request needs an account-only tool, the result lists the available keyless tools and two ways to continue:

{
  "code": "KEYLESS_TOOL_NOT_AVAILABLE",
  "auth_mode": "keyless",
  "available_tools": [
    "firecrawl_scrape",
    "firecrawl_search",
    "firecrawl_parse"
  ],
  "next_actions": [
    {
      "kind": "connect_account",
      "connect_url": "https://firecrawl.dev/connect/mcp"
    },
    {
      "kind": "configure_api_key",
      "header": "Authorization: Bearer <FIRECRAWL_API_KEY>"
    }
  ]
}

To continue, connect through https://mcp.firecrawl.dev/v2/mcp-oauth or add an API key as a Bearer header on the existing /v2/mcp connection, then retry the request.

Installation

You can use the hosted endpoints above or run the server locally. Create an API key at firecrawl.dev/app/api-keys when you need the full tool set or higher limits.

Running with npx

env FIRECRAWL_API_KEY=fc-YOUR_API_KEY npx -y firecrawl-mcp

Manual Installation

npm install -g firecrawl-mcp

Running on Cursor

Install the hosted keyless server in one click (no API key required):

Or add it to ~/.cursor/mcp.json manually:

{
  "mcpServers": {
    "firecrawl": {
      "url": "https://mcp.firecrawl.dev/v2/mcp"
    }
  }
}

Running locally with an API key

Note: Requires Cursor version 0.45.6+ For the most up-to-date configuration instructions, please refer to the official Cursor documentation on configuring MCP servers: Cursor MCP Server Configuration Guide To configure Firecrawl MCP in Cursor v0.48.6

Open Cursor Settings
Go to Features > MCP Servers
Click ”+ Add new global MCP server”

Enter the following code:

{
  "mcpServers": {
    "firecrawl-mcp": {
      "command": "npx",
      "args": ["-y", "firecrawl-mcp"],
      "env": {
        "FIRECRAWL_API_KEY": "YOUR-API-KEY"
      }
    }
  }
}

To configure Firecrawl MCP in Cursor v0.45.6

Open Cursor Settings
Go to Features > MCP Servers
Click ”+ Add New MCP Server”
Enter the following:
- Name: “firecrawl-mcp” (or your preferred name)
- Type: “command”
- Command: env FIRECRAWL_API_KEY=your-api-key npx -y firecrawl-mcp

If you are using Windows and are running into issues, try cmd /c "set FIRECRAWL_API_KEY=your-api-key && npx -y firecrawl-mcp"

Replace your-api-key with your Firecrawl API key. If you don’t have one yet, you can create an account and get it from https://www.firecrawl.dev/app/api-keys After adding, refresh the MCP server list to see the new tools. Open Composer with Command+L (Mac), select Agent, enable Firecrawl MCP, and describe the web data you need.

Running on Windsurf

Add this to your ./codeium/windsurf/model_config.json:

{
  "mcpServers": {
    "mcp-server-firecrawl": {
      "command": "npx",
      "args": ["-y", "firecrawl-mcp"],
      "env": {
        "FIRECRAWL_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

Running with Streamable HTTP Mode

To run the server using streamable HTTP transport locally instead of the default stdio transport:

env HTTP_STREAMABLE_SERVER=true FIRECRAWL_API_KEY=fc-YOUR_API_KEY npx -y firecrawl-mcp

Use the url: http://localhost:3000/v2/mcp locally, or the hosted https://mcp.firecrawl.dev/v2/mcp

Installing via Smithery (Legacy)

To install Firecrawl for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install @mendableai/mcp-server-firecrawl --client claude

Running on VS Code

For one-click installation, click one of the install buttons below…

For manual installation, add the following JSON block to your User Settings (JSON) file in VS Code. You can do this by pressing Ctrl + Shift + P and typing Preferences: Open User Settings (JSON).

{
  "mcp": {
    "inputs": [
      {
        "type": "promptString",
        "id": "apiKey",
        "description": "Firecrawl API Key",
        "password": true
      }
    ],
    "servers": {
      "firecrawl": {
        "command": "npx",
        "args": ["-y", "firecrawl-mcp"],
        "env": {
          "FIRECRAWL_API_KEY": "${input:apiKey}"
        }
      }
    }
  }
}

Optionally, you can add it to a file called .vscode/mcp.json in your workspace. This will allow you to share the configuration with others:

{
  "inputs": [
    {
      "type": "promptString",
      "id": "apiKey",
      "description": "Firecrawl API Key",
      "password": true
    }
  ],
  "servers": {
    "firecrawl": {
      "command": "npx",
      "args": ["-y", "firecrawl-mcp"],
      "env": {
        "FIRECRAWL_API_KEY": "${input:apiKey}"
      }
    }
  }
}

Note: Some users have reported issues when adding the MCP server to VS Code due to how it validates JSON with an outdated schema format (microsoft/vscode#155379). This affects several MCP tools, including Firecrawl. Workaround: Disable JSON validation in VS Code to allow the MCP server to load properly.
See reference: directus/directus#25906 (comment). The MCP server still works fine when invoked via other extensions, but the issue occurs specifically when registering it directly in the MCP server list. We plan to add guidance once VS Code updates their schema validation.

Running on Claude Desktop

Add this to the Claude config file:

{
  "mcpServers": {
    "firecrawl": {
      "url": "https://mcp.firecrawl.dev/v2/mcp"
    }
  }
}

This starts with the keyless tools and keeps an API key out of the Desktop configuration. After the account-endpoint rollout is live, use the browser OAuth flow above for the full account-connected tool set. If you get a “Couldn’t reach the MCP server” error, your Claude Desktop version may not support streamable HTTP transport. Use the local npx approach instead (requires Node.js):

{
  "mcpServers": {
    "firecrawl": {
      "command": "npx",
      "args": ["-y", "firecrawl-mcp"],
      "env": {
        "FIRECRAWL_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

If you see a spawn npx ENOENT error, Node.js is not installed or not in your system PATH. Install Node.js from nodejs.org (LTS version), then fully restart Claude Desktop. On Windows, you can also run where npx in Command Prompt and use the full path (e.g. C:\\Program Files\\nodejs\\npx.cmd) as the command value.

Running on Claude Code

Start keyless, or use the Firecrawl CLI to install an environment-backed Bearer header for the full tool set:

# Remote hosted URL (recommended, works without an API key)
claude mcp add --transport http firecrawl https://mcp.firecrawl.dev/v2/mcp

# Full account limits without storing the key in the URL
export FIRECRAWL_API_KEY=fc-your-api-key
firecrawl setup mcp --agent claude-code --global --yes

Running on Codex

Start keyless:

codex mcp add firecrawl --url https://mcp.firecrawl.dev/v2/mcp

For the full tool set, keep the key in an environment variable and tell Codex which variable supplies the Bearer token:

export FIRECRAWL_API_KEY=fc-your-api-key
codex mcp add firecrawl \
  --url https://mcp.firecrawl.dev/v2/mcp \
  --bearer-token-env-var FIRECRAWL_API_KEY

Running on Google Antigravity

Google Antigravity allows you to configure MCP servers directly through its Agent interface.

Open the Agent sidebar in the Editor or the Agent Manager view
Click the ”…” (More Actions) menu and select MCP Servers
Select View raw config to open your local mcp_config.json file
Add the following configuration:

{
  "mcpServers": {
    "firecrawl": {
      "command": "npx",
      "args": ["-y", "firecrawl-mcp"],
      "env": {
        "FIRECRAWL_API_KEY": "YOUR_FIRECRAWL_API_KEY"
      }
    }
  }
}

Save the file and click Refresh in the Antigravity MCP interface to see the new tools

Replace YOUR_FIRECRAWL_API_KEY with your API key from https://firecrawl.dev/app/api-keys.

Running on n8n

To connect the Firecrawl MCP server in n8n:

Get your Firecrawl API key from https://firecrawl.dev/app/api-keys
In your n8n workflow, add an AI Agent node
In the AI Agent configuration, add a new Tool
Select MCP Client Tool as the tool type
Enter the MCP server Endpoint:

https://mcp.firecrawl.dev/v2/mcp

Set Server Transport to HTTP Streamable
Set Authentication to Bearer and paste your Firecrawl API key, or leave it as None to use the keyless free tier
For Tools to include, select from the tools returned by the server. Keyless mode returns only Search, Scrape, and Parse; authenticated mode returns the full tool set.

For self-hosted deployments, run the MCP server with npx and enable HTTP transport mode:

env HTTP_STREAMABLE_SERVER=true \
    FIRECRAWL_API_KEY=fc-YOUR_API_KEY \
    FIRECRAWL_API_URL=YOUR_FIRECRAWL_INSTANCE \
    npx -y firecrawl-mcp

This will start the server on http://localhost:3000/v2/mcp which you can use in your n8n workflow as Endpoint. The HTTP_STREAMABLE_SERVER=true environment variable is required since n8n needs HTTP transport.

Configuration

Environment Variables

Cloud and self-hosted API

FIRECRAWL_API_KEY: Your Firecrawl API key
- Required when using cloud API (default)
- Optional when using self-hosted instance with FIRECRAWL_API_URL
FIRECRAWL_API_URL (Optional): Custom API endpoint for self-hosted instances
- Example: https://firecrawl.your-domain.com
- If not provided, the cloud API will be used (requires API key)

Configuration Examples

For cloud API usage:

export FIRECRAWL_API_KEY=your-api-key

For self-hosted instance:

export FIRECRAWL_API_URL=https://firecrawl.your-domain.com
export FIRECRAWL_API_KEY=your-api-key  # If your instance requires auth

Custom configuration with Claude Desktop

Add this to your claude_desktop_config.json:

{
  "mcpServers": {
    "mcp-server-firecrawl": {
      "command": "npx",
      "args": ["-y", "firecrawl-mcp"],
      "env": {
        "FIRECRAWL_API_KEY": "YOUR_API_KEY_HERE"
      }
    }
  }
}

Hosted MCP vs local MCP

The hosted MCP server is optimized for safe remote use. Some options that are available when running the MCP server locally are narrowed or unavailable remotely:

Hosted keyless mode exposes exactly Search, Scrape, and Parse at both tool discovery and tool-call time, and is rate-limited per IP.
Local-only file reads are available only when you run the MCP server locally.
Webhooks and local file paths should be configured from a local or self-hosted MCP server when the agent needs access to local resources.

For a fixed, read-only MCP surface with search and research tools, see the Search Only MCP Endpoint.

Rate Limiting

Rate limits are enforced by Firecrawl. Use an API key for higher limits and access to the full tool set.

Available Tools

1. Scrape Tool (`firecrawl_scrape`)

Scrape content from a single URL with advanced options.

{
  "name": "firecrawl_scrape",
  "arguments": {
    "url": "https://example.com",
    "formats": ["markdown"],
    "onlyMainContent": true,
    "waitFor": 1000,
    "mobile": false,
    "includeTags": ["article", "main"],
    "excludeTags": ["nav", "footer"],
    "skipTlsVerification": false
  }
}

To redact personally identifiable information, include redactPII in the scrape tool arguments.

{
  "name": "firecrawl_scrape",
  "arguments": {
    "url": "https://example.com/contact",
    "formats": ["markdown"],
    "redactPII": true
  }
}

2. Map Tool (`firecrawl_map`)

Map a website to discover all indexed URLs on the site.

{
  "name": "firecrawl_map",
  "arguments": {
    "url": "https://example.com",
    "search": "blog",
    "sitemap": "include",
    "includeSubdomains": false,
    "limit": 100,
    "ignoreQueryParameters": true
  }
}

Map Tool Options:

url: The base URL of the website to map
search: Optional search term to filter URLs
sitemap: Control sitemap usage - “include”, “skip”, or “only”
includeSubdomains: Whether to include subdomains in the mapping
limit: Maximum number of URLs to return
ignoreQueryParameters: Whether to ignore query parameters when mapping

Best for: Discovering URLs on a website before deciding what to scrape; finding specific sections of a website. Returns: Array of URLs found on the site.

3. Search Tool (`firecrawl_search`)

Search the web and optionally extract content from search results.

{
  "name": "firecrawl_search",
  "arguments": {
    "query": "your search query",
    "limit": 5,
    "location": "United States",
    "tbs": "qdr:m",
    "scrapeOptions": {
      "formats": ["markdown"],
      "onlyMainContent": true
    }
  }
}

Search Tool Options:

query: The search query string (required)
limit: Maximum number of results to return
location: Geographic location for search results
tbs: Time-based search filter (e.g., qdr:d for past day, qdr:w for past week, qdr:m for past month)
filter: Additional search filter
sources: Array of source types to search (web, images, news)
scrapeOptions: Options for scraping search result pages
enterprise: Array of enterprise options (default, anon, zdr)

4. Parse Tool (`firecrawl_parse`)

Parse a local file such as a PDF, DOCX, XLSX, or HTML document into clean, LLM-ready data.

{
  "name": "firecrawl_parse",
  "arguments": {
    "filePath": "/absolute/path/to/report.pdf",
    "formats": ["markdown"]
  }
}

When you run Firecrawl MCP locally against a Firecrawl API instance with FIRECRAWL_API_URL, the MCP server can read filePath directly and sends the file bytes to /v2/parse. When you use the remote hosted MCP server, the hosted server cannot read files from your machine. In that case firecrawl_parse uses a two-step handoff that also works on the remote keyless URL:

Call firecrawl_parse with filePath. The tool returns a pre-filled upload command and a nextToolCall containing an uploadRef.
Run the upload command on the machine that can read the file, then call firecrawl_parse again with the returned uploadRef.

The upload command sends the file bytes to a short-lived signed upload target. It does not include your Firecrawl API key.

Parse Tool Options:

filePath: Local path to the file you want to parse. Use this for the first call.
uploadRef: Reference returned by the first hosted-MCP call. Use this for the second call after the upload succeeds.
formats: Output formats. Defaults to markdown.
parsers: Parser controls, such as PDF parsing options.
contentType: Optional file MIME type override.
declaredSizeBytes: Optional file size hint. Files are limited to 50 MB.

Best for: Local or non-public documents that are not available at a public URL. Not recommended for: Public document URLs. Use firecrawl_scrape instead; it will detect and parse documents from URLs.

5. Crawl Tool (`firecrawl_crawl`)

Start an asynchronous crawl with advanced options.

{
  "name": "firecrawl_crawl",
  "arguments": {
    "url": "https://example.com",
    "maxDiscoveryDepth": 2,
    "limit": 100,
    "allowExternalLinks": false,
    "deduplicateSimilarURLs": true
  }
}

6. Check Crawl Status (`firecrawl_check_crawl_status`)

Check the status of a crawl job.

{
  "name": "firecrawl_check_crawl_status",
  "arguments": {
    "id": "550e8400-e29b-41d4-a716-446655440000"
  }
}

Returns: Status and progress of the crawl job, including results if available.

7. Extract Tool (`firecrawl_extract`)

Extract structured information from web pages using LLM capabilities. Supports both cloud AI and self-hosted LLM extraction.

{
  "name": "firecrawl_extract",
  "arguments": {
    "urls": ["https://example.com/page1", "https://example.com/page2"],
    "prompt": "Extract product information including name, price, and description",
    "schema": {
      "type": "object",
      "properties": {
        "name": { "type": "string" },
        "price": { "type": "number" },
        "description": { "type": "string" }
      },
      "required": ["name", "price"]
    },
    "allowExternalLinks": false,
    "enableWebSearch": false,
    "includeSubdomains": false
  }
}

Example response:

{
  "content": [
    {
      "type": "text",
      "text": {
        "name": "Example Product",
        "price": 99.99,
        "description": "This is an example product description"
      }
    }
  ],
  "isError": false
}

Extract Tool Options:

urls: Array of URLs to extract information from
prompt: Custom prompt for the LLM extraction
schema: JSON schema for structured data extraction
allowExternalLinks: Allow extraction from external links
enableWebSearch: Enable web search for additional context
includeSubdomains: Include subdomains in extraction

When using a self-hosted instance, the extraction will use your configured LLM. For cloud API, it uses Firecrawl’s managed LLM service.

8. Agent Tool (`firecrawl_agent`)

Autonomous web research agent that independently browses the internet, searches for information, navigates through pages, and extracts structured data based on your query. This runs asynchronously — it returns a job ID immediately, and you poll firecrawl_agent_status to check when complete and retrieve results.

{
  "name": "firecrawl_agent",
  "arguments": {
    "prompt": "Find the top 5 AI startups founded in 2024 and their funding amounts",
    "schema": {
      "type": "object",
      "properties": {
        "startups": {
          "type": "array",
          "items": {
            "type": "object",
            "properties": {
              "name": { "type": "string" },
              "funding": { "type": "string" },
              "founded": { "type": "string" }
            }
          }
        }
      }
    }
  }
}

You can also provide specific URLs for the agent to focus on:

{
  "name": "firecrawl_agent",
  "arguments": {
    "urls": ["https://docs.firecrawl.dev", "https://firecrawl.dev/pricing"],
    "prompt": "Compare the features and pricing information from these pages"
  }
}

Agent Tool Options:

prompt: Natural language description of the data you want (required, max 10,000 characters)
urls: Optional array of URLs to focus the agent on specific pages
schema: Optional JSON schema for structured output

Best for: Complex research tasks where you don’t know the exact URLs; multi-source data gathering; finding information scattered across the web; extracting data from JavaScript-heavy SPAs that fail with regular scrape. Returns: Job ID for status checking. Use firecrawl_agent_status to poll for results.

9. Check Agent Status (`firecrawl_agent_status`)

Check the status of an agent job and retrieve results when complete. Poll every 15-30 seconds and keep polling for at least 2-3 minutes before considering the request failed.

{
  "name": "firecrawl_agent_status",
  "arguments": {
    "id": "550e8400-e29b-41d4-a716-446655440000"
  }
}

Agent Status Options:

id: The agent job ID returned by firecrawl_agent (required)

Possible statuses:

processing: Agent is still researching — keep polling
completed: Research finished — response includes the extracted data
failed: An error occurred

Returns: Status, progress, and results (if completed) of the agent job.

10. Interact with a Page (`firecrawl_interact`)

Interact with a page in a live browser session: click buttons, fill forms, extract dynamic content, or navigate deeper. Use one of two targeting modes:

Pass url to open and interact with a fresh page in one MCP call.
Pass scrapeId from a previous firecrawl_scrape call to reuse the already-loaded page.

Do not pass both url and scrapeId. Provide either prompt or code. scrapeOptions can only be used with url mode. URL mode example:

{
  "name": "firecrawl_interact",
  "arguments": {
    "url": "https://example.com/products",
    "prompt": "Click on the first product and tell me its price"
  }
}

Scrape reuse example:

{
  "name": "firecrawl_interact",
  "arguments": {
    "scrapeId": "scrape-id-from-previous-scrape",
    "prompt": "Click the Sign In button"
  }
}

Interact Tool Options:

url: Page to interact with; opens the session for you. Use this or scrapeId.
scrapeId: Scrape job ID from a previous firecrawl_scrape call. Use this or url.
prompt: Natural language instruction describing the action to take. Provide prompt or code.
code: Code to execute in the browser session. Provide code or prompt.
language: bash, python, or node (optional, defaults to node, only used with code).
timeout: Execution timeout in seconds, 1–300 (optional, defaults to 30).
scrapeOptions: Optional scrape controls used only with url mode.

Best for: Multi-step workflows on a single page — searching a site, clicking through results, filling forms, extracting data that requires interaction. Returns: Interaction result including output and live view URLs.

11. Stop Interact Session (`firecrawl_interact_stop`)

Stop an interact session for a scraped page. Call this when you are done interacting to free resources.

{
  "name": "firecrawl_interact_stop",
  "arguments": {
    "scrapeId": "scrape-id-from-previous-scrape"
  }
}

Interact Stop Options:

scrapeId: The scrape ID for the session to stop (required)

Returns: Confirmation that the session has been stopped.

Logging System

The server includes comprehensive logging:

Operation status and progress
Performance metrics
Credit usage monitoring
Rate limit tracking
Error conditions

Example log messages:

[INFO] Firecrawl MCP Server initialized successfully
[INFO] Starting scrape for URL: https://example.com
[INFO] Starting crawl for URL: https://example.com
[WARNING] Credit usage has reached warning threshold
[ERROR] Rate limit exceeded, retrying in 2s...

Error Handling

The server provides robust error handling:

Structured keyless recovery actions for account-gated tools
Automatic retries for transient errors
Rate limit handling with backoff
Detailed error messages
Credit usage warnings
Network resilience

Example error response:

{
  "content": [
    {
      "type": "text",
      "text": "Error: Rate limit exceeded. Retrying in 2 seconds..."
    }
  ],
  "isError": true
}

For hosted keyless mode, inspect the structured result’s code and next_actions. Use connect_account only when a person can complete a browser flow. Use configure_api_key for CI, servers, and unattended agents. Anonymous callers are never routed into an authentication challenge from this endpoint.

Development

# Install dependencies
npm install

# Build
npm run build

# Run tests
npm test

Contributing

Fork the repository
Create your feature branch
Run tests: npm test
Submit a pull request

Thanks to contributors

Thanks to @vrknetha, @cawstudios for the initial implementation! Thanks to MCP.so and Klavis AI for hosting and @gstarwd, @xiangkaiz and @zihaolin96 for integrating our server.

License

MIT License - see LICENSE file for details

​Features

​Choose how you want to start

Try it now

Connect your account

Run unattended

​Try it now

​Connect your account

​Run unattended

​Upgrade from keyless mode

​Installation

​Running with npx

​Manual Installation

​Running on Cursor

​Running locally with an API key

​Running on Windsurf

​Running with Streamable HTTP Mode

​Installing via Smithery (Legacy)

​Running on VS Code

​Running on Claude Desktop

​Running on Claude Code

​Running on Codex

​Running on Google Antigravity

​Running on n8n

​Configuration

​Environment Variables

​Cloud and self-hosted API

​Configuration Examples

​Custom configuration with Claude Desktop

​Hosted MCP vs local MCP

​Rate Limiting

​Available Tools

​1. Scrape Tool (firecrawl_scrape)

​2. Map Tool (firecrawl_map)

​Map Tool Options:

​3. Search Tool (firecrawl_search)

​Search Tool Options:

​4. Parse Tool (firecrawl_parse)

​Parse Tool Options:

​5. Crawl Tool (firecrawl_crawl)

​6. Check Crawl Status (firecrawl_check_crawl_status)

​7. Extract Tool (firecrawl_extract)

​Extract Tool Options:

​8. Agent Tool (firecrawl_agent)

​Agent Tool Options:

​9. Check Agent Status (firecrawl_agent_status)

​Agent Status Options:

​10. Interact with a Page (firecrawl_interact)

​Interact Tool Options:

​11. Stop Interact Session (firecrawl_interact_stop)

​Interact Stop Options:

​Logging System

​Error Handling

​Development

​Contributing

​Thanks to contributors

​License

Features

Choose how you want to start

Try it now

Connect your account

Run unattended

Upgrade from keyless mode

Installation

Running with npx

Manual Installation

Running on Cursor

Running locally with an API key

Running on Windsurf

Running with Streamable HTTP Mode

Installing via Smithery (Legacy)

Running on VS Code

Running on Claude Desktop

Running on Claude Code

Running on Codex

Running on Google Antigravity

Running on n8n

Configuration

Environment Variables

Cloud and self-hosted API

Configuration Examples

Custom configuration with Claude Desktop

Hosted MCP vs local MCP

Rate Limiting

Available Tools

1. Scrape Tool (`firecrawl_scrape`)

2. Map Tool (`firecrawl_map`)

Map Tool Options:

3. Search Tool (`firecrawl_search`)

Search Tool Options:

4. Parse Tool (`firecrawl_parse`)

Parse Tool Options:

5. Crawl Tool (`firecrawl_crawl`)

6. Check Crawl Status (`firecrawl_check_crawl_status`)

7. Extract Tool (`firecrawl_extract`)

Extract Tool Options:

8. Agent Tool (`firecrawl_agent`)

Agent Tool Options:

9. Check Agent Status (`firecrawl_agent_status`)

Agent Status Options:

10. Interact with a Page (`firecrawl_interact`)

Interact Tool Options:

11. Stop Interact Session (`firecrawl_interact_stop`)

Interact Stop Options:

Logging System

Error Handling

Development

Contributing

Thanks to contributors

License