What is Semantic Scholar?
Semantic Scholar is a free, AI-powered research tool for scientific literature, run by the Allen Institute for AI (Ai2). It lets students, researchers and developers search more than 200 million papers from every field of science, read them with an augmented reader, save them to a library and set alerts, and query the data through a free API.
Keyword search in general engines buries relevant papers under noise and paywalled listings. This tool is built only for scholarly literature: it indexes papers, authors, venues and citations as a connected graph, so you can follow who cites whom and find related work quickly. The same data is open to developers through the Academic Graph API and downloadable datasets, which makes it a common source for literature-review tools and research projects.
Main Features
- Search across more than 238 million papers from all fields of science by title, author, concept or topic.
- Semantic Reader, a beta augmented reading view, adds context to papers as you read them.
- A personal library lets you save and organize papers you want to keep.
- Alerts notify you about new research that matches papers or authors you follow.
- Citation and reference lists show how a paper connects to earlier and later work.
- The Academic Graph API returns data on papers, authors, citations, venues and embeddings, with an optional free API key for higher rate limits.
- A Recommendations API suggests similar papers, and bulk datasets such as S2AG and S2ORC are available for download.
How to Get Started with Semantic Scholar
- Go to semanticscholar.org and type an author, concept or paper title into the search bar.
- Filter results and open a paper to see its abstract, citations and references.
- Create a free account to save papers to your library and turn on alerts.
- Developers can request an API key and start with the Academic Graph documentation.
Semantic Scholar Pricing
The website, library, alerts and API are free. The API works without a key at shared rate limits, and a free key gives an introductory limit of 1 request per second on all endpoints.
Who Should Use It?
- Graduate students building a literature review
- Academic researchers tracking new papers in their field
- Journalists and writers who need primary sources
- Developers and data scientists who need citation data through an API
- Librarians helping patrons find scholarly work
Semantic Scholar Alternatives
- Reflect: better for linking your own reading notes, not for finding papers
- Hex: better for analyzing exported paper data with SQL and Python
- Airtable: better for tracking a shared reading list as a database
Frequently Asked Questions
Is Semantic Scholar free?
Yes. Searching, reading, saving papers to a library and setting alerts are all free, and it is run by the Allen Institute for AI, a nonprofit. The API is also free, with an optional key for higher rate limits.
Does Semantic Scholar have an API?
Yes. It offers the Academic Graph API for papers, authors, citations and venues, a Recommendations API, and downloadable datasets such as S2AG and S2ORC. Unauthenticated calls share a public rate limit, and a free API key starts at 1 request per second.
How many papers does it cover?
The homepage lists more than 238 million papers from all fields of science, and the API page cites about 2.49 billion citations and 79 million authors. Coverage includes computer science, biomedicine, physics, social sciences and more.
