We are stoked to announce this feature!
You can now upload a reference image to guide the model about the spatial awareness, outlines, and overall structure of your desired illustration. You can use this to augment the text prompt for better prompt adherence.
Choose the intensity level on the slider to decide how strongly you want the model to condition its generation based on your image.
Got no reference image but a rough idea in your mind? Fret not! You can simply draw the rough drawing and see it appear in the illustration style of your choosing!
