GPT-4 multimodal input

Few people can already test the GPT-4 with multimodal input. I would be curious to know how the image input can be specified for the GPT-4 API request. Could someone paste a cURL or any request example here?

I’m looking for answers from people who have actually tried it with API calls (and not on the chat interface).

Thank you!

Image input at GPT4 API may be still at beta status. You may like to check at Open AI document.

I know. But I’m looking for answers from people who have actually tried it with API calls (and not on the chat interface).

Sorry I can not be of help to your question.