I want to do something similar to what chagpt plus users can do with the “Make it More” trend (The ChatGPT 'Make It More' Trend Makes Wild AI Images | Lifehacker) where users generate images, then keep editing on them. In this case the model accepts a reference image + a prompt.
Is this possible through the dalle3/GPT4 API? I couldn’t find such information in the documentation.
I have been trying to make a script where I want to push reference image into Dalle 3 API + a prompt generated with GPT 4 vision but as of now its not yet possible to send Image Reference over API. I’m waiting for it as well.