Firecrawl

Search, scrape, and retrieve current web data through Firecrawl’s hosted MCP server, without running a local container. Firecrawl operates and updates the server directly.

See the official Firecrawl MCP documentation for setup, tools, limits, and authentication details.

Authentication method: Enter v2/mcp-oauth for browser sign-in, or v2/mcp to provide an optional API key or use the limited keyless tool set.

Features

• Search and Scrape: Find relevant pages and extract clean, structured content
• Site Discovery and Crawling: Map websites and collect content across multiple pages
• Browser Interaction: Click, navigate, and interact with dynamic pages
• Document Parsing: Parse PDFs, documents, spreadsheets, and HTML files
• Structured Extraction: Extract specific fields using prompts and JSON schemas
• Research: Run autonomous research and search paper, GitHub, and developer indexes
• Monitoring: Track pages and sites for meaningful changes over time

What you’ll need to connect

Choose OAuth to sign in to Firecrawl when prompted, or choose API Key / Keyless to connect without a browser. API-key connections use your Firecrawl team’s plan. Keyless access is limited to Search, Scrape, and Parse.

Provide an API key from https://www.firecrawl.dev/app/api-keys for the full API-key tool surface. Leave the key blank when using OAuth or keyless access.

⚠️ Upgrade notice

If you’re upgrading this server, note that this replaces the Obot-hosted Firecrawl server with Firecrawl’s official hosted MCP server.

• Reconnect required: Select OAuth or API-key/keyless access. The previous required API-key configuration is no longer used.
• Authentication: OAuth and limited keyless access are now available in addition to API-key access.
• Tool changes: Tool names, parameters, and the available scraping, research, interaction, and monitoring capabilities change.