search

PublicAffairs / openai-gemini

Gemini ➜ OpenAI API proxy. Serverless!
https://my-openai-gemini-demo.vercel.app/
MIT License
382 stars 238 forks source link

please support ask from files (images,media,pdf) #16

Closed bugbounted closed 2 months ago

bugbounted commented 2 months ago

hi PublicAffairs please support ask from files (images,media,pdf) & put sample in readme.md waiting for your repsonse regards

johnd0e commented 2 months ago

Hi.

  1. This projects contains no original api, so you'll need to search for OpenAI API examples. As mentioned here in openai-gemini README, only the chat/completions endpoint is supported: for the purposes openai-gemini I can only support those API parts common to both OpenAI and Gemini.
  2. Currently in the api there is direct support for sending images only. There is no direct way for processing PDFs, though it is definitely possible at the application level (so the app may need to extract text or image data on its own).
  3. This project is solely for API translation. You might be looking for a different type of software. There are many applications for interacting with documents. I'm not sure of your exact needs, but here's a random example: https://github.com/haruiz/geminiplayground. You can also search for "RAG" keyword.