Semantic Scholar
Visit Semantic Scholar →About Semantic Scholar
Semantic Scholar is a free AI-powered academic search engine developed by the Allen Institute for Artificial Intelligence (AI2), a nonprofit research organization founded by Paul Allen in 2014. It indexes over 235 million academic papers across all disciplines and applies AI to make the literature more accessible and discoverable.
The service is actively maintained, with no shutdown notices or paywall changes reported. AI2 continues to fund and develop it as a public good for the scientific community.
Current AI capabilities include TLDR summaries (AI-generated one-sentence abstracts for papers), Semantic Reader (beta) with skimming highlights that identify Goal, Method, and Result sections on most English-language arXiv computer science papers, Citation Cards, personalized in-line citations, definitions on demand, and integration with Hypothesis for annotation. Researchers and developers can access the underlying data through the Academic Graph API (S2AG).
Pricing: Semantic Scholar is completely free — the web interface, all AI features, and the S2AG API have no cost, and there are no premium tiers. The API's public endpoints are rate-limited to 1000 requests per second shared among all unauthenticated users; free API keys are available, but the introductory authenticated rate limit is just 1 request per second, with higher limits available on request.
Limitations: Semantic Scholar is a search and discovery tool — it does not extract structured data from papers the way Elicit does, nor does it classify citations as supporting or contrasting the way Scite does. Coverage across disciplines is broad but depth varies, and skimming highlights are currently limited mostly to arXiv computer science papers. Developers building on the API should account for its shared rate limits.
Best suited for researchers, students, and developers who need a free, comprehensive academic search engine with semantic search capabilities and API access — particularly as a complement to Elicit or Scite for specific research workflows.
- Completely free — no cost for any feature including the API
- 235M+ papers across all academic disciplines
- AI TLDR summaries for papers without abstracts
- Semantic Reader with skimming highlights for arXiv CS papers
- Free S2AG API for building research tools and applications
- Discovery tool only — does not extract structured data like Elicit
- No citation relationship classification like Scite (supporting vs. contrasting)
- Coverage depth varies across disciplines — specialized fields may be incomplete
- API rate limits are restrictive: 1000 req/s shared among unauthenticated users, 1 req/s for introductory authenticated keys
- No dedicated customer support — community-supported tool
Choose Semantic Scholar if…
- ✅ You want free, broad academic search across 235+ million papers from any field
- ✅ You want AI-generated TLDR summaries and Semantic Reader's skimming highlights
- ✅ You search for papers by topic without knowing specific citation relationships
- ✅ You want open access PDF links and semantic search that understands research concepts
Leave a Review
Reviews are published after moderation. We don't share your email.