Running local models on Macs gets faster with Ollama's MLX supportarstechnica.com2 points·by kinoriw·há 3 meses·0 comments