Running local models on Macs gets faster with Ollama's MLX support

(arstechnica.com) by kinoriw | Apr 1, 2026 | 0 comments on HN