Ideal inputs
Upload one image with the unwanted object visible and describe its location, size, and the background texture that should continue behind it.
Remove a distracting object from an image by describing the covered area and asking the surrounding scene to be reconstructed naturally.
Upload one image with the unwanted object visible and describe its location, size, and the background texture that should continue behind it.
Use the clearest source and make sure the object to remove is visible rather than already cropped away.
Describe its position and the nearby surfaces, lines, or patterns the edit should continue.
Check seams, repeated textures, shadows, reflections, and any subject parts near the removed area.
The edit workflow generates a filled scene where the described distraction has been removed as far as the source allows.
Large objects, complex perspective, reflections, and areas covering important structure may be reconstructed incorrectly. There is no dedicated mask editor here.
You can describe multiple objects, but a focused request for one area at a time is easier to review and refine.
No. The page uses an uploaded image and a text direction, so describe the target area clearly in the prompt.