Using 'Convert PDF to Word' API on a Hebrew Searchable PDF document gets corrupted

Hello,

I have noticed a strange behavior using Adobe PDF services API. Seems like a bug to me.

I am using Microsoft Power automate with Adobe API in order to process PDF documents in Hebrew.

Using Adobe PDF services API – 'Create Searchable PDF using OCR' works fine in Hebrew.
If I manually copy the Hebrew text from the output searchable PDF document, it pastes correctly into a Word document (in Hebrew).

However, if I use the – 'Convert PDF to Word' API on the searchable PDF document, the Word documents receives corrupted text (Gibberish).

It is important to note that if I use the 'Convert PDF to Word' API on a PDF document that was NOT originally produced through the 'Create Searchable PDF using OCR', it does works fine in Hebrew.

I will appreciate if someone could point me in the right direction solving this issue.