Mathemagic's ComfyUI Workflows [Z-Image Turbo]
Catgirl because why not.
This post publishes my Z-Image Turbo workflows. For other model families, see:
Qwen workflows, including Qwen-Image and Qwen-Image-Edit
Chroma1-HD workflows
Illustrious workflows
Tools: background remover workflow
The workflows are embedded in the images. The text on the sign says attempts to say (as well as the model could, across a few rerolls) which workflow is embedded in the image.
For more details, see my guide.
Notes
The txt2img is the best render out of five. Each time, the model insisted spelling it "Tubro".
The model does get "img2img" right, while Qwen doesn't.
The term "inpaint" invariably became "inpant".
Anyway, "Tubro" and "inpant" are too funny to fix, especially taken together with her enthusiasm.
It boggles the mind that image-genAIs don't know image-genAI vocabulary, although these terms have been standard at least ever since A1111 came out, so off the top of my head, at least since 2023.