Swift 🧨Diffusers - Fast Stable Diffusion for Mac

Transform your text into stunning images with ease using Diffusers for Mac, a native app powered by state-of-the-art diffusion models. It leverages a bouquet of SoTA Text-to-Image models contributed by the community to the Hugging Face Hub, and converted to Core ML for blazingly fast performance. Our latest version, 1.1, is now available on the Mac App Store with significant performance upgrades and user-friendly interface tweaks. It's a solid foundation for future feature updates. Plus, the app is fully open source with a permissive license, so you can build on it too! Check out our GitHub repository at https://github.com/huggingface/swift-coreml-diffusers for more information.

What exactly is 🧨Diffusers for Mac anyway?

The Diffusers app (App Store, source code) is the Mac counterpart to our 🧨diffusers library. This library is written in Python with PyTorch, and uses a modular design to train and run diffusion models. It supports many different models and tasks, and is highly configurable and well optimized. It runs on Mac, too, using PyTorch's mps accelerator, which is an alternative to cuda on Apple Silicon.

Why would you want to run a native Mac app then? There are many reasons:

Performance Benchmarks

TL;DR: Depending on your computer Text-to-Image Generation can be up to twice as fast on Diffusers 1.1. ⚡️

We've done a lot of testing on several Macs to determine the best combinations of compute devices that yield optimum performance. For some computers it's best to use the GPU, while others work better when the Neural Engine, or ANE, is engaged.

Come check out our benchmarks. All the combinations use the CPU in addition to either the GPU or the ANE.

Community Call for Benchmark Data

Other Improvements in Version 1.1

Version 1.1 also includes additional information about what the different generation settings do. We want 🧨Diffusers for Mac to make image generation as approachable as possible to all Mac users, not just technologists.