Question

You are building a lightweight vision app in Microsoft Foundry that sends a customer's uploaded product photo to a multimodal model for interpretation. During development, you need to decide how to pass the image to the model's chat completion request. Which approach correctly supplies the image so the multimodal model can interpret it?

Accepted Answer

Include the image in the message content as a URL reference or a base64-encoded data URI alongside the text prompt.

Answer

Convert the image to plain text using OCR first, then send only the extracted text to the model.

Answer

Upload the image to a separate image-classification endpoint because chat models cannot accept images.

Answer

Attach the image as a code interpreter file so the model can run Python to open it.