Apple released a set of optimizations to Core ML to enable running the Stable Diffusion text-to-image model on Apple Silicon-powered devices running the latest iOS or macOS versions, respectively iOS 16.2 and macOS 13.1.