OpenGraph.io

What Is an MCP Server? Give Your AI Web Superpowers

An MCP Server (Model Context Protocol server) securely connects AI assistants to web data. Extract metadata, capture screenshots, scrape pages, and query content — all from your AI workflow.

No credit card required.

MCP Server Architecture
AI AssistantClaude, Cursor, VS Code...
JSON-RPC
opengraph-io-mcpruns locally on your machine
REST API
OpenGraph.io APIapi.opengraph.io
Supported AI Clients
Claude
Cursor
VS Code
Windsurf
JetBrains
Zed
Cline
+ More
npx -y opengraph-io-mcp
Why MCP

MCP Server Benefits: Why Use the OpenGraph.io MCP Server?

The OpenGraph.io MCP server is a secure local bridge that gives your AI assistant real-time access to metadata extraction, screenshot capture, web scraping, and AI-powered page queries — working across 7+ AI clients through the standard AI agent protocol, with no custom integration code required.

Metadata Extraction via AI

Your AI assistant can extract OpenGraph tags, Twitter Cards, and HTML meta tags from any URL — no code required.

MCP metadata extraction

AI-Powered Screenshots

Capture full-page screenshots at any viewport size, directly from your AI workflow. Ideal for competitive analysis and QA.

MCP screenshot capture

Intelligent Web Scraping

Scrape rendered HTML with JavaScript execution and bot detection bypass. Get the full page content your AI needs.

MCP web scraper

Ask Questions About Any Page

The AI Query tool lets your assistant ask natural-language questions about webpage content and get structured answers.

MCP AI query
Capabilities

MCP AI Integration: Tools & Capabilities

Seven powerful tools your AI assistant discovers and calls automatically through the Model Context Protocol — the open AI agent protocol for tool access.

Web Data Tools

Get OG Data
get_og_data

Extract OpenGraph metadata, Twitter Cards, and HTML meta tags from any URL.

/api/1.1/site
Scrape
scrape

Fetch raw HTML with JavaScript rendering and bot detection bypass.

/api/1.1/scrape
Screenshot
screenshot

Capture full-page screenshots at configurable viewport sizes.

/api/1.1/screenshot
Query
query

Ask natural-language questions about any webpage and get structured answers.

/api/1.1/query

Image Generation Tools

Generate Image
generate_image

Create AI-generated illustrations, diagrams, icons, and social cards.

AI Image Generation
Iterate Image
iterate_image

Refine and modify previously generated images with follow-up prompts.

AI Image Generation
Export Image
export_image

Save generated images directly to your local project filesystem.

Local Filesystem
How It Works

How the OpenGraph.io MCP Server Works

Set up in under two minutes. Your AI assistant automatically discovers the available tools and calls them when relevant — no code changes needed.

  • 1Configure your AI client: Add the OpenGraph.io MCP config with a one-line npx command and your App ID.
  • 2AI discovers available tools: Your assistant automatically sees metadata extraction, scraping, screenshot, and query tools.
  • 3Ask naturally, get results: Just ask your AI to “get the OG data for this URL” or “screenshot this page” — it calls the right MCP tool.
MCP Server Flow
AI requests: “Get OG data for this URL”
MCP Server calls OpenGraph.io API
Structured JSON returned to AI
Security

MCP Server Security & Architecture for Enterprises

Enterprise-grade security built in. Benefits of MCP servers for enterprises: your credentials never leave your machine and the AI model never sees your API key.

API Key Stays Local

Your App ID is stored in a local config file. It's never transmitted to the AI model or any third party.

Local Execution

The MCP server runs as a subprocess on your machine. No cloud hosting, no shared infrastructure.

Stdio Transport

Communication uses standard input/output — not network requests. No open ports, no attack surface.

No Persistent State

The server doesn't store request data, cache responses locally, or maintain any persistent session state.

Use Cases

MCP Server Use Cases for Developers

Whether you're auditing SEO, analyzing competitors, or building AI-assisted tools — the OpenGraph.io MCP server fits right into your workflow.

  • SEO & AEO workflows: AI agents extract metadata for real-time SEO audits and Answer Engine Optimization insights.
  • Competitive analysis: Screenshot and scrape competitor pages directly from your IDE without switching context.
  • Content migration: Extract and transform web content with AI assistance — headings, paragraphs, and structured data.
  • Link preview generation: Build rich link unfurling in AI-assisted chat apps, bots, and messaging platforms.
  • Automated QA: Query page content to validate deployments, check OG tags, and verify structured data is correct.

SEO Metadata Audit

Extract and validate OG tags, Twitter Cards, and meta descriptions across your entire site.

Competitor Screenshots

Capture competitor pages at any viewport size without leaving your editor.

Content Scraping & Migration

Pull page content with full JS rendering for migration, analysis, or transformation.

Comparison

MCP Server vs. Direct API Integration

Traditional API integrations require custom code. An MCP server lets your AI handle it naturally.

Direct API Integration

  • Write custom HTTP request code for each endpoint
  • Manage authentication tokens and API key rotation
  • Parse and transform JSON responses manually
  • Build error handling, retries, and rate limiting logic
  • Context-switch between IDE and browser to test
  • Maintain integration code as APIs evolve

OpenGraph.io MCP Server

  • AI calls the right tool automatically — zero integration code
  • API key stays local in your config file, never sent to AI
  • Works across 7+ AI clients (Claude, Cursor, VS Code, etc.)
  • Real-time results delivered directly in your AI workflow
  • One-line setup with npx — get started in under 2 minutes
  • MCP server auto-updates via npx — no maintenance needed
Setup

Set Up in Under 2 Minutes

Select your AI client, paste the config, and your assistant gains instant access to web data tools.

claude_desktop_config.json
{
  "mcpServers": {
    "opengraph": {
      "command": "npx",
      "args": [
        "-y",
        "opengraph-io-mcp"
      ],
      "env": {
        "APP_ID": "YOUR_OPENGRAPH_APP_ID"
      }
    }
  }
}

Requires Node.js 18+ for npx. Get your free App ID at dashboard.opengraph.io. See the full MCP documentation or view pricing plans.

Our Clients

Trusted by Thousands

Over 1B Site Previews!

Try it for free today

Get started with 100 free API requests per month. No credit card required. Experience the power of our API with zero commitment.

FAQ

An MCP server (Model Context Protocol server) is a local bridge that securely connects AI assistants — like Claude, Cursor, and VS Code Copilot — to external tools and data sources. The OpenGraph.io MCP server exposes web scraping, metadata extraction, screenshot capture, and AI-powered page querying tools so your AI assistant can access real-time web data without custom integration code.
A traditional API requires you to write HTTP requests, handle authentication, and parse responses manually in your code. An MCP server wraps those same API capabilities and exposes them as tools that AI assistants can discover and call automatically — no integration code required. The AI decides when to call the right tool based on your natural language request.
The OpenGraph.io MCP server works with Claude Desktop, Cursor, VS Code (Copilot Chat), Windsurf, JetBrains AI Assistant, Zed, and Cline. Any client that supports the Model Context Protocol standard can connect to it.
Yes. The MCP server can extract OpenGraph tags, Twitter Card metadata, and HTML meta tags from any publicly accessible URL. It handles JavaScript-rendered pages, bot detection bypass, and returns structured JSON — all through a single tool call from your AI assistant.
The MCP server runs locally on your machine as a subprocess of your AI client. Your OpenGraph.io App ID is stored in a local configuration file and is never transmitted to the AI model. Communication between the AI client and the MCP server uses stdio (standard input/output), not network requests.
Absolutely. With the OpenGraph.io MCP server, AI agents can extract metadata for SEO audits, capture competitor page screenshots, scrape page content for analysis, and query pages to validate structured data — all from within your AI-assisted workflow. This enables real-time, AI-driven SEO and Answer Engine Optimization insights.
The MCP server itself is free and open-source. It connects to the OpenGraph.io API, which offers a free tier with 100 credits per month. Paid plans start at $25/month for 50,000 credits. Metadata extraction costs 1 credit per request, screenshots cost 20 credits (1 if cached), and AI Query costs 100–200 credits depending on the model. See the full pricing page at https://www.opengraph.io/pricing for details.

Still have Questions? 🤔

Can't find the answer you're looking for? Please chat to our friendly team. We usually respond within 24 hours.