Description

Mima is a simple but powerful utility that makes use of Apple's open-source Stable Diffusion framework to generate images from text.

• Designed for brainstorming and/or visualisation workflows by allowing easy queueing up of items, drag and drop, easy tweaking of generated results, and to stay out of your way while it works. It is not meant to be a deep and involved AI image generation solution, but a lightweight utility that does one thing well.

• Supports cloning of existing images based on the provided text, with configurable strength.

• Supports activating the Stable Diffusion safety filter to minimise chances of running into explicit or disturbing imagery.

• The default engine is Stable Diffusion v1.5 but it can be changed to v2.0, v2.1, XL or even 1.4 if preferred.

• It's free, since I primarily created it for myself after wrangling with a lot of browser-based interfaces and really wanted a nice clean native app. That said, feel free to leave a tip in the "about" window if you like it :)

What's New in Version 1.1.4

• Configurable interval between preview updates
• Clearer appearance for images generated via Stable Diffusion XL
• Updated to the latest ML SDK for performance and stability