Need help understanding the best way to process SlideShare PDF files with the OpenAI API

Documentation
,
1

Hi everyone,

I am reading the OpenAI documentation but still little confused.

I have some study presentation PDFs downloaded using a SlideShare Downloader. My goal is to use the OpenAI API to summarize them, extract key points, and ask questions about the content.

What is the recommended workflow now?

Should I extract all the text first and send it to the Responses API, or is there a better way to work with PDF documents directly?

I checked the documentation, but I’m not sure which approach is the current recommended one. If someone has a simple example or can point me to the right documentation page, it would really help.

Thanks in advance.

2

Use a completions or responses call directly with the base64 encoded file contents:

Here’s a PR to my Discourse Chatbot with that exact feature implementation:

3

Have you looked at the examples in the OpenAI Cookbook?

Search for PDF; several relevant examples are available, including approaches for parsing PDFs and using them with retrieval or file search.

If you have not explored the Cookbook before, it is worth taking some time to browse through it. It contains many useful examples, suggestions, and implementation patterns.

One caution: some older examples may use APIs, models, or techniques that have since been superseded, so check the publication date and compare the code with the current OpenAI documentation before adopting it.

4

Thanks! I didn’t know about the Cookbook. I’ll check the PDF examples here: Cookbook and compare them with the current documentation. My PDF files are from SlideShare Downloader, so I wanted to make sure I’m following the recommended workflow before building anything. Thanks for pointing me in the right direction!