Wallpyria — AI Model & Open-Source Licenses

The optional Frame Expansion (AI) feature of Wallpyria downloads AI models (≈6.6 GB) that run entirely on your Mac. Your images are never uploaded. The models and open-source components used by this feature, together with their licenses, are listed below. Full license texts are linked from each entry and are also distributed alongside the model files.

AI models (downloaded on demand)

Stable Diffusion XL — Inpainting UNet, VAE, Text Encoders (Core ML conversions)
Copyright © 2023 Stability AI · License: CreativeML Open RAIL++-M (July 26, 2023)
Derived from stabilityai/stable-diffusion-xl-base-1.0, diffusers/stable-diffusion-xl-1.0-inpainting-0.1 and apple/coreml-stable-diffusion-xl-base, converted to Core ML format for on-device execution.
Real-ESRGAN ×2 super-resolution (Core ML conversion)
Copyright © 2021 Xintao Wang · License: BSD 3-Clause
CLIP BPE tokenizer (vocab.json / merges.txt)
Copyright © 2021 OpenAI · License: MIT

Open-source code included in the app

PNDM scheduler (from apple/ml-stable-diffusion 1.1.1)
Copyright © 2022–2024 Apple Inc. · License: MIT

Use restrictions notice (CreativeML Open RAIL++-M)

The Stable Diffusion XL models are provided under the CreativeML Open RAIL++-M license, which includes use-based restrictions (Attachment A of the license). By downloading the models and using the Frame Expansion feature you agree not to use the models or generated content for the restricted uses listed in the full license text — including, among others, unlawful purposes, harming or exploiting minors, generating or disseminating false information with intent to harm, defaming or harassing others, or generating personal identifiable information with intent to harm. The full, binding list is set out in Attachment A of the license.