Draw Things

Draw Things provides a local interface for running Stable Diffusion and related generative models on iOS, iPadOS, and macOS devices. Users can enter textual prompts to generate images, apply style transformations, and create short video clips from generated frames, all without requiring an internet connection. The app also offers an “Infinite Canvas” that lets users expand the workspace and edit any region of an image, and it supports on‑device LoRA training for custom model fine‑tuning.

The software is aimed at creators who want a private, offline workflow for AI‑assisted image and video production. It bundles model downloading, prompt‑based generation, style transfer, and LoRA training into a single application, allowing artists, designers, and hobbyists to experiment and iterate quickly on their own hardware. A free edition supplies basic local generation, while optional paid tiers unlock higher cloud compute limits and additional features such as custom LoRA uploads.

Key distinguishing aspects include the ability to run the full Stable Diffusion pipeline locally on Apple devices, the integration of image‑to‑video and infinite‑canvas tools, and the option to train LoRAs entirely on‑device, preserving user privacy while providing a complete creative workflow in one app.