macOS 27 ships a free local LLM on Apple Silicon Macs. I made it easy to use from Node and Python

Excerpt

Apple Silicon Macs on macOS 26+ come with a small LLM built in. No download, no API key, and nothing leaves your Mac. Why I built it I was making a tool that writes API docs from code, and I didn't want users to install Ollama or paste an API key. Apple's model was already on their Mac, so I used it. Getting it to work well was harder than expected. At temperature 0 it kept repeating itself, and a 2-second call took 20. Calls took 17 seconds instead of 1.5 until I kept one process running. So I