Hi @keyes ,
With the first URL, I currently get the following error:
We are sorry, we were not able to process this. (Input content was too short)
Upon further investigation, it appears to be because the PDF has no embedded text - it is a bunch of scanned images. We currently do not do OCR, so we can't support summarizing these kinds of PDFs at the moment.
The second URL is because we have temporarily suspended audio extraction capabilities - we are rolling out an update soon to improve the error message here.