Can this model do text2image?
As the title says, can I use it like flux.1-dev to generate images from only text input?
I tried replace the flux.1-dev model file with this one, and kept the rest of workflow unchanged, but the result is very bad.
The technical report mentioned that the dev version only focuses on image-to-image.
I on the other hand had bad results using it for image-to-image, but good results using it for text-to-image (replacing accordingly flux1-dev and flux-fill in my workflows)
The technical report mentioned that the dev version only focuses on image-to-image.
That could be the case. I read elsewhere that the max and pro version generates nice images.
As the title says, can I use it like flux.1-dev to generate images from only text input?
yes it can text to image and image to image, but text to image generation is weak