Yes, it lazy-loads papers using credentials from institutional subscribers and then stores them for fast retrieval. When you're the first person to request a pa

Yes, it lazy-loads papers using credentials from institutional subscribers and then stores them for fast retrieval. When you're the first person to request a paper you'll see an animation of a biphenyl molecule rotating while the site fetches the paper from an institution with access. Sometimes, there is no corresponding institution with access currently hooked up to sci-hub and the retrieval process fails. But I've only seen that happen a few times.