
Decodo MCP Server
Scrape websites, search engines, eCommerce, and social media for AI agents via Decodo's Web Scraping API.
Add to your client
Copy the config for your MCP client and paste it into its config file.
npx -y @decodo/mcp-serverPaste into ~/Library/Application Support/Claude/claude_desktop_config.json
{
"mcpServers": {
"decodo-mcp-server": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://mcp.decodo.com/mcp",
"--header",
"Authorization: Basic <basic_auth_token>"
]
}
}
}Claude Desktop connects to remote servers through the `mcp-remote` proxy (installed on first run via npx). Restart Claude Desktop after saving.
Step-by-step guides: Add to Claude Desktop · Add to Cursor · Add to Windsurf
Before you start
- A Decodo account with a Web Scraping API Basic authentication token (free tier: up to 2,000 requests, no credit card required)
- An MCP-compatible client such as Claude Desktop, Cursor, or Windsurf
- Node.js 18.0+ (required for the local/npx stdio setup)
About Decodo MCP Server
Decodo MCP Server is a web scraping layer for AI agents that connects MCP-compatible clients to Decodo's Web Scraping API. It enables web scraping for LLMs, real-time data retrieval for RAG, AI agent browsing and research, and structured data extraction from dynamic websites — without maintaining proxies, parsers, or retry logic. Tools are grouped into toolsets (web, search, ecommerce, social_media, ai), each enabling targeted scraping of general web pages, search engines (Google, Bing), marketplaces (Amazon, Walmart, Target, TikTok Shop), social platforms (Reddit, TikTok, YouTube), and AI search engines (ChatGPT, Perplexity). Outputs are available as Markdown, JSON, or screenshots, with optional geo-targeting, JavaScript rendering, device emulation, and token-limit truncation inferred from natural-language prompts.
Tools & capabilities (30)
scrape_as_markdownScrapes any target URL given via prompt and returns the results in Markdown.
screenshotCaptures a screenshot of any webpage and returns it as a PNG image.
google_searchScrapes Google Search for a given query and returns parsed results.
google_adsScrapes Google Ads search results.
google_lensScrapes Google Lens image search results.
google_ai_modeScrapes Google AI Mode (Search with AI) results.
google_travel_hotelsScrapes Google Travel Hotels search results.
amazon_searchScrapes Amazon Search for a given query and returns parsed results.
amazon_productScrapes an Amazon product page.
amazon_pricingScrapes Amazon product pricing information.
amazon_sellersScrapes Amazon seller information.
amazon_bestsellersScrapes an Amazon Bestsellers list.
walmart_searchScrapes Walmart Search for a given query and returns parsed results.
walmart_productScrapes a Walmart product page.
target_searchScrapes Target Search for a given query and returns parsed results.
target_productScrapes a Target product page.
tiktok_postScrapes a TikTok post URL for structured data (e.g. engagement, caption, hashtags).
tiktok_shop_searchScrapes TikTok Shop Search for a given query and returns parsed results.
tiktok_shop_productScrapes a TikTok Shop product page.
tiktok_shop_urlScrapes a TikTok Shop page by URL.
youtube_metadataScrapes YouTube video metadata.
youtube_channelScrapes YouTube channel videos.
youtube_subtitlesScrapes YouTube video subtitles.
youtube_searchSearches YouTube videos.
reddit_postScrapes a specific Reddit post.
reddit_subredditScrapes Reddit subreddit results.
reddit_userScrapes a Reddit user profile and their posts/comments.
bing_searchScrapes Bing Search results.
chatgptSearches and interacts with ChatGPT for AI-powered responses and conversations.
perplexitySearches and interacts with Perplexity for AI-powered responses and conversations.
What this server can do
Decodo MCP Server provides tools for these capabilities — tap one to see every MCP server that does the same:
When to use it
- AI-powered web scraping — give LLMs the ability to collect fresh data instead of relying on static training data.
- RAG with live data — pull real-time Google, Bing, and AI search results into retrieval pipelines.
- eCommerce intelligence — track product prices, listings, and sellers across marketplaces without getting blocked.
- Social media data collection — gather posts, channels, and engagement data from Reddit, TikTok, and YouTube.
- Travel and price aggregation — build tools that collect live pricing and availability across websites.
Security notes
Authentication uses a Decodo Web Scraping API Basic auth token. For the hosted server pass it as an 'Authorization: Basic <basic_auth_token>' header; for the local stdio server pass it via the SCRAPER_API_TOKEN environment variable. Keep this token secret — do not commit it to source control.
Decodo MCP Server FAQ
Do I need to build my own scraping infrastructure?
No. Decodo handles proxy rotation, CAPTCHA solving, anti-bot systems, and JavaScript rendering server-side, giving you a single integration point for reliable web data.
How do I authenticate?
Use a Web Scraping API Basic authentication token from the Decodo dashboard. For the hosted server, pass it as an 'Authorization: Basic <basic_auth_token>' header; for the local stdio server, set it via the SCRAPER_API_TOKEN environment variable.
Can I enable only some of the tools?
Yes. Tools are grouped into toolsets (web, search, ecommerce, social_media, ai). Pass a comma-separated list via the TOOLSETS env var (local) or the ?toolsets= query parameter (hosted URL). When none are specified, all tools are registered.
Can I run it locally instead of using the hosted server?
Yes. Clone the repo, run npm install and npm run build, then point your MCP client at the built build/index.js with 'command': 'node' — or use the published @decodo/mcp-server package via npx. Node.js 18+ is required.
Is there a free tier?
Yes. Creating a free Decodo account gives you up to 2,000 free requests with no credit card required.
Alternatives to Decodo MCP Server
Compare all alternatives →Official MCP reference server that fetches a URL and returns its content as clean Markdown, with chunking.
Official Firecrawl MCP server — scrape, crawl, map, search, and structured extraction for any LLM client.
Exa's neural web search and crawling MCP server — runs locally via npx or as a hosted remote endpoint.
Compare Decodo MCP Server with: