How to manipulate images using image and text as input and get image output

I have images that i want to manipulate. when using ChatGPT it works (by providing a link to the updated image), but when using python with openai I only get text output.
How to get the updated image programmatically?
(I tried to use “client.images.edit” with “dall-e-2” and the results are not good and not the same as ChatGPT)…