CSL styles take a locale variable that's not based on the language of the source, but on the language specified for the bibliography. E.g. if I'm writing in English, all my et als, ands, editors, etc. are going to be in English, even for Hebrew, German, French, Chinese, etc. sources.

It looks like your implementation of CSL defaults to English. You could change this to Hebrew (we don't know how you've implemented CSL, so can't help much with that) and you'd get ו and אחר
for all items.

Hi
Thank you for you comment, we do provide the locale information- passing he-IL when recognizing Hebrew records and we still get back from the in the response the conjunctions & and et-al instead of the Hebrew ones. we are using the citeproc API , can you recommend on a way to debug this outside of our application to see if there are issues in the conjunctions handling on non-Latin languages

Any input will help us with improving the use of CSL for our Israeli customers
All the best
Nili

this isn't an issue with CSL in general -- e.g. I get correct conjunctions when passing the he-IL locale to Zotero -- so you'll really need to take this up with your citeproc implementation. The first step for debugging would be to figure out the exact stack you're using for Primo's CSL implementation. There are e.g. at least half a dozen different citeprocs and "citeproc API" isn't something that I recognize/know about under that label.