ktransformers

GitHub

A Flexible Framework for Experiencing Heterogeneous LLM Inference/Fine-tune Optimizations

19,211 stars Python 1 file ยท ~11,727 tokens

ktransformers MCP Server

MCP server for kvcache-ai/ktransformers

Ready-to-Use Client Configurations
{
  "mcpServers": {
    "ktransformers": {
      "command": "npx",
      "args": [
        "-y",
        "@kvcache-ai/ktransformers"
      ]
    }
  }
}