Mooncake Model Context Protocol (MCP) Server

Mooncake is the serving platform for Kimi, a leading LLM service provided by Moonshot AI.

6,210 stars

C++ 1 file · ~6,530 tokens #disaggregation #inference #kvcache #llm #rdma #reinforcement-learning #sglang #tokenspeed

Mooncake MCP Server

MCP server for kvcache-ai/Mooncake

Ready-to-Use Client Configurations

{
  "mcpServers": {
    "Mooncake": {
      "command": "npx",
      "args": [
        "-y",
        "@kvcache-ai/Mooncake"
      ]
    }
  }
}

{
  "mcpServers": {
    "Mooncake": {
      "command": "npx",
      "args": [
        "-y",
        "@kvcache-ai/Mooncake"
      ]
    }
  }
}

{
  "mcpServers": {
    "Mooncake": {
      "command": "npx",
      "args": [
        "-y",
        "@kvcache-ai/Mooncake"
      ]
    }
  }
}

{
  "mcpServers": {
    "Mooncake": {
      "command": "npx",
      "args": [
        "-y",
        "@kvcache-ai/Mooncake"
      ],
      "disabled": false,
      "autoApprove": []
    }
  }
}

{
  "mcpServers": {
    "Mooncake": {
      "command": "npx",
      "args": [
        "-y",
        "@kvcache-ai/Mooncake"
      ]
    }
  }
}

docker run -i --rm mcp/Mooncake