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