MCP Directory
Back

vmlx

by jjang-ai · Python · ★ 725

vMLX - JANGTQ Uber Compressed MLX Models - L2 Disk Cache (survives restart) + L1 Paged (super fast ttft) + Hybrid SSM Scheduler + Cont Batching + etc!

#anthropic-api#kvcache-compression#kvcache-optimization#kvcache-reuse#llm#lmstudio#macbook#mcp-server#mlx#mlxllm#mlxstudio#omlx#omlx-alternative#openai-api#openclaw#openclaw-agent#persistent-memory#prefix-cache#vmlx

Install

pip install git+https://github.com/jjang-ai/vmlx.git

Auto-generated from the repo's primary language — check the repo's own README for its exact install instructions.

Claude Desktop config

A starting point for your claude_desktop_config.json (auto-generated — the repo's README may specify a published package name or required env vars):

{
  "mcpServers": {
    "vmlx": {
      "command": "uvx",
      "args": [
        "git+https://github.com/jjang-ai/vmlx.git"
      ]
    }
  }
}

From the README

MLX Inference Server for Apple Silicon Self-hosted inference server for LLMs, VLMs, and image generation on Apple Silicon. OpenAI + Anthropic + Ollama compatible HTTP API. Self-hosted; no third-party API keys required. Native MTP artifact detection and family-specific cache policy gates keep speculative/cache settings explicit and model-safe. Looking for a native Swift macOS app or Swift inference engine? See osaurus.ai. Quickstart • Models • Features • Image Gen • API • Desktop App • JANG • CLI • Config…
Read full README on GitHub →

Related ai & ml servers