Semantic Scholar, built by the nonprofit Allen Institute for AI (AI2) and launched in 2015, is a free, large-scale academic search and citation index covering more than 200 million papers, notable in this category less as a standalone competitor and more as underlying infrastructure: several other tools covered in this directory, including Afforai/Logically, Epsilon, and Paperguide, explicitly draw on Semantic Scholar's dataset or API as one of their own source layers.
That infrastructure role matters for understanding the category as a whole: choosing a polished AI research assistant usually means choosing an interface layered on top of a shared underlying dataset like Semantic Scholar's, rather than a genuinely different paper database from one competitor to the next. Used directly, Semantic Scholar itself includes citation graphs, influential-citation highlighting, and AI-generated TL;DR summaries per paper.
Semantic Scholar is completely free to use, funded by the nonprofit Allen Institute for AI rather than a subscription model. For a researcher who wants direct, no-cost access to the same large academic dataset several paid AI research assistants are built on top of, without a polished chat interface layered in between, Semantic Scholar's free, direct-access model addresses that specific need.
