How can I get DALL-E to reference the files I uploaded into my assistant?

Ok, so I am playing around with the assistant and the API. Bot is pretty straight forward… ask the user some questions, then, once I have collected what I need use a custom function to call DALL-E to make an image based on the collected data.

My client would like to upload images of their previous work so that the DALL-E images look more like what they make… but I am not sure how to do that. Uploading the files into the assistant won’t help right? Because DALL-E doesn’t know about it.

Is there something I am missing with the DALL-E API?

What you are asking for is not possible.

Well bummer. So I can upload files to the assistant to use and reference, but as soon as I go to the world of DALL-E all bets are off eh?

Again, bummer.

assistants is just a somewhat enhanced chat/completions.

Dall-E is an entirely different model and model architecture. It doesn’t have (at least not publicly exposed) any capacity for ingesting images for reference.

Two completely distinct models with distinct abilities.

You can ask the ChatGPT Vision model to describe an image and use that as input to the Dall-E model, but that will almost certainly not get you the results you’re looking for.