Paperless ngx seems to support fulltext search of PDF’s, in addition to other document types.

I use Karakeep which uses Mellisearch as a part of its stack

It’s part of Bar Assistant stack. Haven’t got into it outside of it

Yep. I have an instance I self-host to power a “cookbooks -> scanner -> images -> OCR -> indexed search” tool I built for myself. The whole thing is pretty much just a few bash scripts thinly wired into a Meilisearch instance, but the utility is helpful when looking for a recipe from my collection.