I'm using OpenAI API (based on langchain) to do a task of summarizing serials of PDF files into a single PDF file.
By using pdfreader
I can parse and construct pdf files, and by gpt I can easily get summarized text.
The questions is, the source pdf files contains images, some of them are crucial for understanding the overall content, thus I want the summarized content from gpt is not just text, but with the key images from original pdf files, does gpt support this ?