Remote Model-Context-Protocol (MCP) server that lets LLMs run Globalping network measurements (ping, traceroute, DNS, HTTP, etc.) through natural-language.
https://github.com/jsdelivr/globalping-mcp-serverStop switching between AI assistants and network debugging tools. The Globalping MCP Server bridges that gap by giving Claude, GPT, and other LLMs direct access to real-time network measurements from thousands of locations worldwide.
You're troubleshooting a connectivity issue, performance bottleneck, or CDN behavior. The usual workflow involves:
This back-and-forth kills your debugging flow and makes it nearly impossible for AI to provide contextual insights about network behavior patterns.
The Globalping MCP Server eliminates this friction by connecting your AI assistant directly to Globalping's global probe network. Your AI can now run network tests from any location and immediately analyze the results within the same conversation.
Real-time global perspective: Instead of testing from your local machine, run measurements from thousands of probes across different continents, ISPs, and cloud providers.
Context-aware debugging: Your AI maintains full context of measurement results, enabling intelligent follow-up questions and comparative analysis across locations.
Natural language interface: Skip memorizing command syntax - just ask "Is example.com faster from Europe or Asia?" and get comprehensive analysis.
CDN Performance Analysis
"Test page loading times for our app from major cities in Asia, Europe, and North America. Which regions are seeing degraded performance?"
Your AI runs HTTP tests from multiple locations simultaneously and provides actionable insights about geographic performance patterns.
DNS Propagation Verification
"Check if our DNS changes have propagated globally - test both A and CNAME records from different continents"
Get immediate verification across global locations without manually checking multiple DNS tools.
Connectivity Troubleshooting
"Users in China can't reach our API. Run traceroutes from Beijing, Shanghai, and Guangzhou to see where packets are getting dropped"
Your AI identifies the exact network hops where connectivity fails and suggests potential solutions.
Network Path Analysis
"Compare routing paths to our servers from AWS regions vs Google Cloud regions. Which offers better latency?"
Make infrastructure decisions based on real routing data rather than theoretical performance claims.
The server runs remotely at https://mcp.globalping.dev/sse - no local installation needed.
Claude Desktop: Add this to your config:
{
"mcpServers": {
"globalping": {
"command": "npx",
"args": ["mcp-remote", "https://mcp.globalping.dev/sse"]
}
}
}
Cursor: Add to your MCP configuration and restart.
Anthropic Console: Add as a custom tool when creating assistants.
Works with any MCP-compatible AI client - the protocol handles all the complexity.
Location targeting supports continent codes (EU, AS, NA), country codes (US, DE, JP), cities (London, Tokyo), cloud regions (aws-us-east-1), and even specific networks (Cloudflare, Google).
Traditional network debugging tools give you data. The Globalping MCP Server gives you intelligent analysis. Your AI can spot patterns across multiple measurements, correlate issues with geographic regions, and suggest targeted solutions based on comprehensive network data.
Instead of interpreting raw traceroute output yourself, get insights like "Latency spikes consistently at hop 7 for European users - this suggests issues with the London-Paris peering connection."
Perfect for developers working on distributed systems, SREs managing global infrastructure, or anyone who needs to understand network behavior beyond their local environment.
Time to stop babysitting network tools and let your AI handle the heavy lifting.