Local AI on the Mac

Local AI on the Mac is a practical, 330-page guide to running large language models on Apple Silicon instead of renting them from the cloud.
It walks through a real M2 Max (64GB) setup: local inference with Ollama and MLX, agent tool-calling that hits the terminal and mail, long-context coder models, on-device speech, and hybrid routing so the machine stays first and the cloud stays optional.
This book is for Mac users who want a private assistant for code, research, and daily work — especially anyone who cannot send client or personal data to a hosted model.
You get the working configs, the model tradeoffs, and the failure modes: RAM pressure, broken tool calls, and how to keep the stack from rotting.