Open ai API can create images. There is a separate api for that called DALLE. The API has two versions DALLE 2 and DALLE 3. Dalle 2 generates much less realistic images, but it supports editing with ai - you simply make 2 copies, and erase some part from one of them and send them both to the api with a prompt. The api then completes the erased parts based on your prompt. But this is DALLE 2, DALLE 3 doesn’t support editing yet. Your use case seems complex. Good luck.
1 Like