I mean, they were shit for a very long period of time. But at least in my domain they've (I think?) gotten much better. I wouldn't cite any Chinese research in my work from 10-20 years go. These days, I really need to scrutinize something. There is still definitely a paper mill aspect to what I read (I was reviewing a paper as a referee the other day and I swore I was missing it, until like, the 5th re-read, and yes. They had no N for their sample size), but like, there definitely has been a shift.