ScrapingBee offers a free tier with starter credits (1,000 on registration), and higher volume requires a paid plan. Your ScrapingBee API key meters credits per request against your account.

Does it support remote/OAuth?

It is remote-only over Streamable HTTP at https://mcp.scrapingbee.com/mcp. Authentication is the API key passed as an api_key query parameter on the URL rather than OAuth or a Bearer header; clients without native remote support can bridge with npx mcp-remote against the keyed URL.

How is the API key passed?

As a query parameter: append ?api_key= to the MCP URL. The server explicitly rejects the key sent as an Authorization Bearer header — it must be in the URL.

ScrapingBee MCP server

OfficialScrapingBeeConfig last verified Jun 1, 2026

ScrapingBee's official MCP server: scrape pages to text or HTML, screenshot, extract data, search the web, and pull Amazon, Walmart, and YouTube data.

ScrapingBee MCP is ScrapingBee's official server that fronts its managed scraping API — proxy rotation, headless browser rendering, and anti-bot handling — as a set of agent tools. An agent can fetch a URL as clean text or raw HTML, capture a full screenshot, download a file, or run AI-style extraction that pulls specific fields off a page. Beyond generic scraping it ships purpose-built tools for the searches teams actually run: a fast web search, structured Google results as a fallback, Amazon and Walmart search and product details, and a full YouTube suite covering search, video metadata, transcripts, and a trainability check. It even exposes an ask_chatgpt tool for a quick model response inside the same server.

ScrapingBee hosts a remote Streamable HTTP endpoint at https://mcp.scrapingbee.com/mcp. Authentication is your ScrapingBee API key passed as an api_key query parameter on the MCP URL (a Bearer header is not accepted), so clients that support remote MCP connect by appending ?api_key=<your key>. For clients without native remote support, ScrapingBee documents bridging with npx mcp-remote against the same keyed URL. A free tier with starter credits is available, and the key meters credits per request.

Quick install

Copy-paste configs are provided for all 8 supported clients. Pick your client below.

Add to ~/.claude.json

~/.claude.json

json

{
  "mcpServers": {
    "scrapingbee": {
      "type": "http",
      "url": "https://mcp.scrapingbee.com/mcp"
    }
  }
}

Or via CLI

bash

claude mcp add --transport http scrapingbee https://mcp.scrapingbee.com/mcp

Step-by-step guides

Available tools

Tool	Description
get_page_text	Scrape a URL with ScrapingBee and return the page content as clean text.
get_page_html	Scrape a URL with ScrapingBee and return the page content as raw HTML.
get_screenshot	Scrape a URL with ScrapingBee and return a screenshot of the page.
get_file	Fetch a file from a URL and return it in a format suitable for the client.
extract_page_data	Scrape a URL and extract specific structured data from the page.
fast_search	Search the web using ScrapingBee's Fast Search API and return structured results.
get_google_search_results	Fallback tool that returns structured Google search results.
ask_chatgpt	Get a response from a ChatGPT model via the ScrapingBee API.
get_amazon_search_results	Scrape Amazon search results, with country, currency, domain, and sort options.
get_amazon_product_details	Scrape Amazon product details for a given ASIN.
get_walmart_search_results	Scrape Walmart search results.
get_walmart_product_details	Scrape Walmart product details.
get_youtube_search_results	Search YouTube and return structured results.
get_youtube_video_metadata	Fetch structured metadata for a YouTube video, including title and description.
get_youtube_video_transcript	Fetch the transcript (captions/subtitles) for a YouTube video.
check_youtube_video_trainability	Check whether a YouTube video has transcripts available and suitable for use.

Required configuration

SCRAPINGBEE_API_KEYRequired
ScrapingBee API key from scrapingbee.com, passed as an api_key query parameter on the MCP URL (not as a header). Required.

What you can do with it

Scrape protected pages without infrastructure

Point the agent at a JavaScript-heavy or bot-protected page and call get_page_text or get_screenshot; ScrapingBee handles proxies, headless rendering, and anti-bot so the agent just gets clean content back.

Pull e-commerce and YouTube data

Use get_amazon_search_results or get_walmart_product_details for retail research, or the YouTube suite to fetch a video's metadata and transcript, all as structured JSON from one server.

FAQ

Is it free?: ScrapingBee offers a free tier with starter credits (1,000 on registration), and higher volume requires a paid plan. Your ScrapingBee API key meters credits per request against your account.
Does it support remote/OAuth?: It is remote-only over Streamable HTTP at https://mcp.scrapingbee.com/mcp. Authentication is the API key passed as an api_key query parameter on the URL rather than OAuth or a Bearer header; clients without native remote support can bridge with npx mcp-remote against the keyed URL.
How is the API key passed?: As a query parameter: append ?api_key=<your key> to the MCP URL. The server explicitly rejects the key sent as an Authorization Bearer header — it must be in the URL.

← Browse all search-and-data servers