chat-ui icon indicating copy to clipboard operation
chat-ui copied to clipboard

Published 20 hours ago • verified publisher icon huggingface

Analyze uploaded PDF files through OpenAI API

Open zu0feng opened this issue 8 months ago • 2 comments

When I upload a PDF file and leverage it, I will get the base64 data. But I didn't find the code to process it in endpoints/openai, while it can handle the image base64 data. Besides, I failed to transfer it back to text. How can I analyze the file through OpenAI API?

Image

zu0feng avatar Mar 09 '25 09:03 zu0feng

What model are you using - I don't think the OpenAI models have native PDF Support do they? I added this for the Anthropic endpoint a couple of months ago.

evalstate avatar Mar 11 '25 08:03 evalstate

I would also like to add audio at the same run. Since Gemini supports both and has an openai api this could make sense. Especially when running a litellm proxy

flozi00 avatar Mar 15 '25 18:03 flozi00