Shypd.aiThe largest open index of AI toolsnative-diffusion enables native Stable Diffusion inference on iOS and macOS devices using MPSGraph. It offers a Swift Package Manager wrapper for Maple Diffusion, adding image-to-image capabilities and a flexible API for local model execution.
Trending
About
native-diffusion is an open-source project that facilitates native Stable Diffusion inference directly on iOS and macOS devices. Unlike many other solutions, it leverages the MPSGraph framework in Swift, eliminating the need for Python environments. This tool provides a Swift Package Manager wrapper for Maple Diffusion, enhancing its functionality with features like image-to-image generation, Combine publishers, and async/await support. It boasts an extremely simple API, allowing users to generate images with a single line of code, while also offering flexibility for custom parameters like prompt, guidance scale, steps, and seed. The package supports downloading weights from local or remote URLs, including app bundles, and includes a conversion script for optimizing .ckpt models into its memory-optimized format, making it suitable for both macOS and iOS devices. It is optimized for SwiftUI and offers fast performance on newer Macs.
Capabilities
Pricing & Plans
Open Source
Free
FAQs
Trending
