Will we have image OUTPUT for api of gpt-4o?

Title. I know we have input image for the api of gpt-4o, but will we have image output eventually?

dalle-3 is still way too expensive for what I want (flashcards image generation) so I’m hoping gpt-4o is cheaper

1 Like

Eventually, hopefully in the coming months.

Here’s several rounds of refining the 0-shot input to have what you want produced by GPT-4o in ChatGPT Plus, with the added benefit that the text is not full of nonsense scribbles.

In the same fashion, you could have the API generate the code for you to perform your task, accepting a list of inputs, making a page-worth of cards, transforming image to PDF, etc. Not everything has to be entirely AI generated.

My idea is to make flashcards for language-learning.
For example, the front of the card contains the Russian word for “dog”, the back of the card is a picture of a dog.
So I really need the AIgen images.