Well it certainly exists as a social category, and if you want to know whether it exists as a biological category you had better ask geneticists, not social scientists!

Race is a really lousy but not uniformly useless way to categorize genetic relatedness. "Black" is a completely pointless category in this regard. The greatest divergence between human genomes all exists within sub-Saharan Africa. "White", on the other hand, is actually pretty much a thing, as is "Asian" and "Native American". The boundaries are still a mess, and when you have people with multiple ancestries, "multi-racial" doesn't get you very far.

Then we can ask: okay, given that "Asian" does kinda reflect some common genetic origin, are there any important consequences of being "Asian" or not? For almost everything the answer is "no, every race contains tons of variability and has a similar average". There are a few instances of disease or gene-related trait (e.g. lactose intolerance) that segregate somewhat with race, but even then, you want to be way more precise.

So: it's a very bad way to classify people even if you only care about relatedness. And then if you care about using it for anything important, it's even worse. But as a genetic thing it doesn't completely "not exist".