Image input¶
Most models take a text prompt only. A model that also accepts an input image says so on its page and in
the table below; passing image to any other model raises an error.
| Model | Image input | Option |
|---|---|---|
joyai-image-edit |
optional | image |
joyai-image-edit-plus |
required | image |
qwen-image-edit |
required | image |
qwen-image-edit-2509 |
required | image |
sdxl-refiner-1.0 |
required | image |
flux1-kontext-dev |
optional | image |
visualcloze-384 |
required | image |
omnigen-v1 |
optional | image |
from PIL import Image
from imagehub import pipeline
with pipeline("t2i", backend="flux1-kontext-dev", model="checkpoints/flux1-kontext-dev") as edit:
result = edit("Add a red hat to the scene", image=Image.open("photo.png"), seed=0)
result.images[0].save("edited.png")
imagetakes aPIL.Image.Imageor animagehub.Image. Pillow is only needed for the first.- When the option is
optional, leaving it out generates from the text alone. - There is no separate "edit" task: the same
t2ipipeline takes the extra option.