Running local models on Macs gets faster with Ollama's MLX supportarstechnica.com2 points·by kinoriw·il y a 3 mois·0 comments