Llm Latency Tracker MCP server
{
"mcpServers": {
"llm-latency-tracker": {
"type": "http",
"url": "https://llmlatency.dev/mcp"
}
}
}{
"mcpServers": {
"llm-latency-tracker": {
"type": "http",
"url": "https://llmlatency.dev/mcp"
}
}
}{
"mcpServers": {
"llm-latency-tracker": {
"type": "http",
"url": "https://llmlatency.dev/mcp"
}
}
}{
"servers": {
"llm-latency-tracker": {
"type": "http",
"url": "https://llmlatency.dev/mcp"
}
}
}{
"mcpServers": {
"llm-latency-tracker": {
"type": "http",
"url": "https://llmlatency.dev/mcp"
}
}
}file .mcp.json (in your project root)file claude_desktop_config.jsonfile ~/.cursor/mcp.jsonfile .vscode/mcp.jsonfile ~/.codeium/windsurf/mcp_config.jsonversion 1.1.0
What it does
Exposes measured latency, time-to-first-token, and uptime data for ~45 AI inference APIs across four global regions (Europe, US, Asia, South America). Built for developers and AI agents who need real, continuously-measured performance metrics rather than provider claims.
- Query live latency rankings and uptime percentages for AI inference providers across specific regions to select the best endpoint for your application.
- Compare time-to-first-token performance across multiple LLM providers before committing to a service.
- Check the deprecation calendar to identify upcoming model retirements and recommended migration targets.
Tools this server exposes
| Tool | Does |
|---|---|
| get_ai_api_latency | Retrieve measured edge latency, inference latency (time-to-first-token), and uptime rankings for AI inference APIs, optionally filtered by region (eu-hetzner, u |
Before you enable it
An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.