Globalping MCP Server

What is Globalping?

Globalping is a free, public API that provides access to a globally distributed network of probes for monitoring, debugging, and benchmarking internet infrastructure. With Globalping, you can run network tests (ping, traceroute, DNS, MTR, HTTP) from thousands of locations worldwide.

What is the Globalping MCP Server?

The Globalping MCP Server implements the Model Context Protocol (MCP), allowing AI models like OpenAI's GPT and Anthropic's Claude to interact with Globalping's network measurement capabilities through natural language.

It also supports two authentication methods: OAuth and API token authentication. Both methods offer a secure way to interact with our API and provide higher rate limits associated with your account.

Key Features

• 🌐 Global Network Access: Run measurements from thousands of probes worldwide
• 🤖 AI-Friendly Interface: Any LLM will easily parse the data and run new measurements as needed
• 📊 Comprehensive Measurements: Support for ping, traceroute, DNS, MTR, and HTTP tests
• 🔍 Smart Context Handling: Provides detailed parameter descriptions for AI clients to intelligently select measurement types and options
• 🔄 Comparative Analysis: Allows to compare network performance between different targets
• 🔑 Authentication Support: Use OAuth or API token with your Globalping account for higher rate limits

Authentication

The Globalping MCP server supports two authentication methods:

• OAuth Authentication: Automatically handled by the server for secure access
• API Token Authentication: Manual token configuration via Authorization header

Both methods provide higher rate limits and priority access to the probe network.

Using Globalping API Token

The server automatically detects when an API token is provided in the Authorization header and uses it for authentication instead of OAuth.

Getting Your API Token

1. Visit dash.globalping.io
2. Sign in to your account
3. Navigate to Tokens to generate a new API token

Configuration with Authentication

Streamable HTTP transport:

{
    "mcpServers": {
        "globalping": {
            "url": "https://mcp.globalping.dev/mcp",
            "headers": {
                "Authorization": "Bearer YOUR_GLOBALPING_API_TOKEN"
            }
        }
    }
}

Legacy SSE transport:

{
    "mcpServers": {
        "globalping": {
            "command": "npx",
            "args": [
                "mcp-remote",
                "https://mcp.globalping.dev/sse",
                "--header",
                "Authorization: Bearer YOUR_GLOBALPING_API_TOKEN"
            ]
        }
    }
}

Connecting AI Assistants

You can use our MCP server with any MCP-compatible AI assistant, including:

• Claude Desktop
• Anthropic Assistants
• Cursor
• Windsurf
• Any custom implementation of the MCP protocol

See your tool's MCP documentation for details on connecting clients to this server.

Available Tools

• ping - Perform a ping test to a target
• traceroute - Perform a traceroute test to a target
• dns - Perform a DNS lookup for a domain
• mtr - Perform an MTR (My Traceroute) test to a target
• http - Perform an HTTP request to a URL
• locations - List all available Globalping probe locations
• limits - Show your current rate limits for the Globalping API
• getMeasurement - Retrieve a previously run measurement by ID
• compareLocations - Guide on how to run comparison measurements
• help - Show a help message with documentation on available tools

Location Specification

Locations can be specified using the "magic" field, which supports various formats:

• Continent codes: "EU", "NA", "AS", etc.
• Country codes: "US", "DE", "JP", etc.
• City names: "London", "Tokyo", "New York", etc.
• Network names: "Cloudflare", "Google", etc.
• ASN numbers: "AS13335", "AS15169", etc.
• Cloud provider regions: "aws-us-east-1", "gcp-us-central1", etc.

You can also combine these with a plus sign for more specific targeting: "London+UK", "Cloudflare+US", etc.

Development

The codebase is organized into modules:

• src/index.ts - Main entry point and MCP agent definition
• src/app.ts - OAuth web routes
• src/api - Globalping API client
• src/auth - Authentication utilities
• src/config - Configuration and constants
• src/lib - Utility functions
• src/mcp - MCP tool handlers
• src/types - TypeScript type definitions
• src/ui - HTML templates

Add Globalping credentials

Add Globalping OAuth credentials:

• npx wrangler secret put GLOBALPING_CLIENT_ID

KV storage

Used for OAuthProvider docs https://github.com/cloudflare/workers-oauth-provider

• create a KV namespace and copy ID
• binding for it must be OAUTH_KV
• configure kv_namespaces in the wrangler.jsonc file

Installation

The remote MCP server is available under these endpoints:

• Streamable HTTP transport: https://mcp.globalping.dev/mcp
• SSE transport: https://mcp.globalping.dev/sse

You can integrate our Globalping MCP server with various AI tools that support the Model Context Protocol.

Here are instructions for the top 3 most popular tools:

Gemini CLI

To add the Globalping MCP server to Gemini CLI:

1. Ensure you have the Gemini CLI installed.
2. Run the following command:

gemini extensions install https://github.com/jsdelivr/globalping-mcp-server

This will automatically configure the server with OAuth authentication enabled. You will be prompted to log in when you first use the extension.

Note: If you prefer to use a Globalping API token (for higher rate limits or automation), you can manually add the server with the Authorization header instead of using the extension installer:

gemini mcp add globalping https://mcp.globalping.dev/mcp --header "Authorization: Bearer YOUR_TOKEN"

Claude Desktop App

Add to your Claude Desktop configuration file (located at %APPDATA%\Claude\config.json on Windows or ~/Library/Application Support/Claude/config.json on macOS):

[!note] Claude Desktop natively supports only stdio transport for local MCP servers. For remote MCP servers, use the mcp-remote bridge.

{
    "mcpServers": {
        "globalping": {
            "command": "npx",
            "args": [
                "mcp-remote",
                "https://mcp.globalping.dev/sse"
            ]
        }
    }
}

Anthropic Claude API (via Console)

When creating a Claude Assistant in the Anthropic Console:

1. Go to console.anthropic.com
2. Navigate to the Assistants section
3. Create a new Assistant or edit an existing one
4. In the Tools section, select "Add custom tool"
5. Enter the following details:
   • Tool Name: Globalping
   • Description: Run network tests from locations worldwide
   • Tool URL: https://mcp.globalping.dev/mcp (Streamable HTTP transport) or https://mcp.globalping.dev/sse (SSE transport)

Cursor

To add the Globalping MCP server to Cursor:

1. Open Cursor settings
2. Navigate to the Tools & MCP tab
3. Click on "+ New MCP server"
4. This opens the mcp.json config file, where you will need to add:

Streamable HTTP transport:

{
    "mcpServers": {
        "globalping": {
            "url": "https://mcp.globalping.dev/mcp"
        }
    }
}

Legacy SSE transport:

{
    "mcpServers": {
        "globalping": {
            "command": "npx",
            "args": [
                "mcp-remote",
                "https://mcp.globalping.dev/sse"
            ]
        }
    }
}

5. Save and restart Cursor

Usage Examples

Once connected to an AI model through a compatible MCP client, you can interact with Globalping using natural language:

Ping google.com from 3 locations in Europe

Run a traceroute to github.com from Japan and compare with traceroute from the US

Check the DNS resolution of example.com using Google DNS (8.8.8.8)

Is jsdelivr.com reachable from China? Test with both ping and HTTP

What's the average response time for cloudflare.com across different continents?