DiffusionBee
An open-source Stable Diffusion app for macOS that generates and edits images offline, with support for SDXL, ControlNet and local model training.
DiffusionBee is an open-source AI art app for Mac users who want to generate and edit images on their own computer. It runs Stable Diffusion offline, with image generation processed on the device. Model downloads require network access, and optional image uploads can send images externally. Its visual interface suits artists and designers who want local image tools without working through code.
Text prompts can create an image or guide changes to an existing one. Generative fill adds or removes objects in selected areas, while inpainting and outpainting let you revise part of a picture or extend it beyond its original edges. The AI canvas combines prompts with drawing for edits that need more direct control.
DiffusionBee also includes upscaling, image variants, illusion generation and tools for AI animation and video. Control images guide the structure of a result. It supports SD 1.x, SD 2.x and SDXL models, along with ControlNet, LoRA and inpainting models. You can download models within the app and revisit results through generation history.
Custom model training runs locally too. Using your own image dataset, you can train models to generate particular people, objects or styles.
The app requires macOS 13.1 or later and recommends Apple Silicon. Intel Macs also work, though generation is much slower without a dedicated graphics chip. The app uses AGPL-3.0; model licenses apply separately. The project identifies CreativeML Open RAIL-M terms for its Stable Diffusion outputs.
