Model Serving

Name: Model Serving
Rating: 4 (3 reviews)
Author: gitwalter

MCP Server

by gitwalter · Added 5mo ago

Claude

Install

git clone https://github.com/gitwalter/cursor-agent-factory

claude mcp add model-serving

Copy & paste ready

~/Library/Application Support/Claude/claude_desktop_config.json (macOS) %APPDATA%\Claude\claude_desktop_config.json (Windows)

{
  "mcpServers": {
    "model-serving": {
      "command": "npx",
      "args": [
        "@modelcontextprotocol/model-serving"
      ]
    }
  }
}

Paste into your Claude Desktop config file, then restart Claude.

Production model serving with vLLM, TGI, Ollama, Triton, FastAPI endpoints, batching, and GPU optimization