Openai-image inpainting example is not working

Documentation
,
2

The prior conclusion that what you’d get is similar to inpainting an existing “image-1/gpt-4o” image in ChatGPT is backed by further evidence like yours.

A sunlit indoor space features a cozy seating area with wicker furniture and a small pool, surrounded by large windows and lush plants. (Captioned by AI)

The gpt-image-1 example, also used in the API reference, shows no need for a mask or a base image to remix several images into a new one.

Unaltered original? dall-e-2 gives near pixel accuracy and with just minor glitches sometimes around the mask area. Grabbing their picture just now.

A sunlit lounge with wicker furniture and a decorative plant next to an indoor pool is displayed on a computer screen with editing options. (Captioned by AI)
A sunlit lounge with wicker furniture and a decorative plant next to an indoor pool is displayed on a computer screen with editing options. (Captioned by AI)1526×1056 180 KB

And with DALL-E 2 just as broken as it has been, for going on a month:

A sunlit solarium features a small pool, wicker furniture, and lush green plants, creating a serene and inviting atmosphere. (Captioned by AI)
A sunlit solarium features a small pool, wicker furniture, and lush green plants, creating a serene and inviting atmosphere. (Captioned by AI)1024×1024 149 KB

Perhaps okay for infill if you didn’t use the polar bear prompt to now produce an unrecognizable blob.

2 Likes