Zotero automatically detects library records, news items, journal articles, and other objects you might like to save in your collection. Zotero uses so called translators to detect and import data from websites. There are currently more than 500 different translators, facilitating data import from thousands of sites.
Zotero imports records from many library cataloging systems, providing seamless import from hundreds of academic and non-academic libraries. Supported library catalogue systems include: Aleph, Amicus, BiblioCommons, Dynix, Encore, Mango, InnoPAC, Primo, SirsiDynix, TLC/YouSeeMore, Voyager, and WorldCat.
Zotero imports data, and in many cases full-text PDFs, from the most popular electronic databases, including EBSCO, IEEEXplore, JSTOR, Google Scholar, ProQuest, PubMed, and many more. It also works with most major journal publishers, including Cambridge University Press, Oxford University Press, Project MUSE, ScienceDirect (Elsevier), SpringerLink, Taylor and Francis, and many more.
Zotero has dedicated translators for hundreds of websites, ranging from Amazon (in various countries), The New York Times, and The Economist, to Mainichi Daily News (Japan), Kommersant (Russia), Spiegel Online (Germany), and many more from around the world.
Zotero detects and imports metadata embedded by an increasing number of websites and databases in open formats such as COinS, Embedded RDF, Google/HighWire meta tags, and unAPI.
You can find a full, automatically generated list of all Zotero translators here. The list is generated daily and includes automatic tests of most translators, so that Zotero developers and volunteers can quickly identify and repair issues. You can see the code for Zotero's translators at the Zotero Translator GitHub repository.