Very well I would say (20gb and counting). More importantly: Zotero is the only one that does Web scraping correctly. If you are looking at a paywalled article, it will capture your DOM from the browser, so you get what you are looking at. DevonThink and many others choke at paywalled content because they try to access the content independently.

Last but not least: thanks to the developers for the latest updates that bring zotero to a more modern experience. The iOS app specifically is awesome!