Scrapfly

Scrapfly MCP Server connects your AI assistants to live web data through the Model Context Protocol. It enables real-time access to any website, automated anti-bot handling, and structured data extraction so your AI can fetch current information, navigate pages, and return clean results on demand.

How to use

You use the Scrapfly MCP Server by pointing your MCP client at the remote MCP endpoint or by running a local MCP process. Your AI can fetch live pages, extract structured data, capture screenshots, and manage usage through authentication. Start by configuring a connection to the remote MCP server, then invoke the available tools to perform web requests, data extraction, or page analysis. The server handles live web access, anti-bot challenges, and browser-like interactions so your AI stays productive without manual scraping steps.

How to install

{
  "servers": {
    "scrapfly-cloud-mcp": {
      "type": "http",
      "url": "https://mcp.scrapfly.io/mcp"
    }
  }
}

Additional setup notes

You can also run a local or self-hosted instance using a CLI configuration that enables a local MCP process. The following configuration runs a local process via NPX to connect to the remote MCP URL.

Configuration and local deployment

Local or self-hosted deployments support two primary wiring patterns: a remote HTTP MCP URL for general use, or a local stdio process that runs a command to start the MCP client. The included examples show how to connect using either approach and how to pass an API key when needed.

Self-hosted / Local Deployment Details

If you choose to run the server locally, use the following command patterns to start the MCP server and specify the API key if required.

Tools and usage patterns

Leverage the available tools to perform common tasks. For example, call the page fetch tool to retrieve a page, then use the scraper tool to extract structured data, or take a screenshot for visual verification.

Security and authentication

Authenticate requests using OAuth2 for production deployments or an API key for development and personal use. Protect your API key and rotate credentials as needed.

Examples of use cases

• Fetch current product prices from a retailer page and return a structured listing. - Retrieve the latest headlines from a news site and extract summaries. - Capture visual snapshots of a page region for monitoring or QA.

Available tools

scraping_instruction_enhanced

Get best practices & POW token for enhanced scraping guidance; recommended as the first call before other actions.

web_get_page

Quick page fetch with smart defaults to retrieve a webpage's content.

web_scrape

Full control with browser automation to scrape pages, including login flows and dynamic content.

screenshot

Capture screenshots of pages or specific elements for visual analysis.

info_account

Check usage & quota to manage your Scrapfly account.