An OpenAI/Anthropic-compatible LLM inference server for Apple Silicon, built on MLX - View it on GitHub