Do you know anything about how they're doing this? Is it special content in the http headers of the requests from Google? Or do they explicitly know which IP-addresses Google will be doing it's web crawling from?
It's just a different version targeted at Googlebot. Google's IPs aren't secret, and Google is open about how to identify (and verify) their webcrawlers: https://support.google.com/webmasters/answer/80553