Model - Unstable Diffusers xl

Prompt - dark side of the moon.
Results
This could be a Pink Floyd copywrite.



Same Model
Prompt - light side of the moon
Results

Prompt - bugs bunny on the dark side of the moon
Model - ICBINP ( I can't believe it's not photography)

Results -

Prompt- draculas coffin
Results -

Rod, I understand text2image and upscale but not image to image or inpainting. Can you explain them to me please?

Inpainting you have to add an alpha channel to an image and make a selection. For instance select the eyes and delete that area ( keep the selection), and use a prompt like -
add glasses

The inpaint model will paint in glasses where you have your selection.

Image to image you open any image and use a prompt like -
mona lisa style painting.

Hope that explains it for you.
Currently text to image can only create 1 image.(in Gimp diffusion for Gimp-3.0), in Gimp-2.10 you can create 10 with text to image.
Inpaint and image to image can create up to 10 i believe.

Thank you

You are very welcome. I do not know if you need an alpha channel for image to image. You will have to test that.