# Sci-Net Peer-to-peer paper exchange launched in 2025 by [[Alexandra Elbakyan]] as a sister platform to [[Sci-Hub]]. When a paper is missing from the Sci-Hub corpus, a researcher can post a request on Sci-Net; another researcher with institutional access through their university library can pull the PDF and share it back. It addresses Sci-Hub's structural weakness. Publishers have tightened automated scraping over the years, so the freshest papers and certain hardened journals never make it into the Sci-Hub corpus. Sci-Net routes around this by re-introducing the human in the loop; a real researcher with valid institutional credentials performs the request manually, which is much harder to detect or block at scale than a scraper. Architecturally it shifts the access model from "scraping" to "social". The corpus stops being a fixed dataset and becomes a network effect; the more researchers participate, the wider the institutional access surface, the higher the fulfillment rate for arbitrary requests. It is the same shift that turned [[BitTorrent]]-style file sharing into a more durable system than centralized hosting; spread the points of failure across the participants. Together, Sci-Hub, Sci-Net, and [[Sci-Bot]] form a three-layer stack; Sci-Hub provides bulk corpus access, Sci-Net handles the long tail and the freshness gap, Sci-Bot adds natural-language synthesis on top. Each layer reinforces the others. ## References - Sci-Net launch context (C&EN, in Sci-Bot coverage): https://cen.acs.org/policy/publishing/Sci-Hub-created-new-AI/104/web/2026/04 - Sci-Hub mirrors (entry points to the broader ecosystem): https://sci-hub.ru/, https://sci-hub.ee/ ## Related - [[Sci-Hub]] - [[Sci-Bot]] - [[Alexandra Elbakyan]]