Paper search and indexes

Every research project starts with a search, and the place you search shapes what you find. Academic search engines and indexes differ in what they cover, how they decide what counts as a scholarly source, and how far you can document and repeat a search later. Knowing those differences saves time and keeps a literature review honest.

Broad engines come first for most people. Google Scholar indexes articles, theses, books and preprints from a huge range of sources and shows who cited each item. Semantic Scholar covers a large share of the scientific literature, adds short machine summaries to many papers, and shows how later work cites a paper. Both are free and fast, and both are uncurated: duplicate versions, predatory journals and unreviewed items show up next to peer-reviewed work.

Subject and subscription indexes sit at the other end. PubMed covers biomedicine and the life sciences with careful subject indexing. Scopus and Web of Science are curated, multidisciplinary citation databases that most universities license, and they are often expected for documented searches in theses and systematic reviews. Dimensions links papers to grants, patents and clinical trials, which helps when you need to see how research is funded or used.

Open infrastructure has grown quickly. OpenAlex is a free catalog of works, authors and institutions that many newer tools are built on. CORE and BASE gather open access papers from repositories around the world, which helps you find a free, legal copy of an article you cannot reach through a paywall. arXiv is the main preprint server for physics, mathematics, computer science and statistics, so new work often appears there months before journal publication.

11 AI research tools in paper search and indexes

Google Scholar

by Google

Free search across scholarly articles, theses, books and court opinions, with "cited by" counts and links to full text where available.

  • Free plan or free tier
  • Shows sources for its answers

Semantic Scholar

by Allen Institute for AI

Free academic search engine from the Allen Institute for AI, with one-line paper summaries, citation context and a public API.

  • Free plan or free tier
  • Developer API
  • Shows sources for its answers

Consensus

Search engine that answers research questions from peer-reviewed papers and links each statement to the study it came from.

  • Free plan or free tier
  • Shows sources for its answers

OpenAlex

by OurResearch

Free, open catalog of scholarly works, authors, institutions and journals, with an API used for bibliometrics and research tools.

  • Free plan or free tier
  • Developer API

CORE

by The Open University and Jisc

Aggregator of open access research papers from repositories and journals worldwide, with full-text search and an API.

  • Free plan or free tier
  • Developer API

BASE

by Bielefeld University Library

Bielefeld Academic Search Engine: a free search across records harvested from thousands of academic repositories.

  • Free plan or free tier

PubMed

by U.S. National Library of Medicine

Free search of the biomedical and life sciences literature from the U.S. National Library of Medicine, with MeSH indexing.

  • Free plan or free tier
  • Developer API

arXiv

by Cornell University

Open preprint server for physics, mathematics, computer science, statistics and related fields, free to read and search.

  • Free plan or free tier
  • Developer API

Dimensions

by Digital Science

Research database linking publications, grants, patents, clinical trials and policy documents, with a free version for searching papers.

  • Free plan or free tier

Scopus

by Elsevier

Subscription abstract and citation database from Elsevier, usually reached through a university library, with an AI search feature.

Web of Science

by Clarivate

Subscription citation index from Clarivate, used for documented searches and citation analysis through institutional access.

Question-led engines such as Consensus take a plain question and answer it from research papers, with each statement linked to a study. They are good for a quick sense of a topic, especially outside your own field. They are most useful as a starting point next to a structured search, since a summary can reflect how a question was phrased as much as what the evidence says.

A practical habit is to run a broad search in Google Scholar or Semantic Scholar, then repeat the key terms in the subject database your field expects, and write down the exact search strings, filters and dates. That record makes your search repeatable and is often required in methods sections. Set up alerts on your main terms so new papers come to you instead of the other way round.

Coverage also differs by discipline and language. Humanities and social science work, books, and research published outside English are thinner in some engines than in others, so a search that looks complete in one tool can miss a whole body of work. If your topic crosses fields, search at least one index from each field.

The tools below are listed in no particular order. Each one has a short description we wrote and a link to its official site. Check whether your library already provides access before you pay for anything, since many of these databases are licensed by universities.

Questions about paper search and indexes

Is Google Scholar enough for a literature review?

For a first look, often yes. For a thesis chapter or a published review, most fields also expect a documented search in a curated subject database such as PubMed, Scopus or Web of Science, because Google Scholar results cannot be exported or repeated as precisely.

How do I get the full text of a paper behind a paywall?

Check your library first, since it may already license the journal. Then look for an open access copy through CORE, BASE or the author's repository page, or ask your library for an interlibrary loan. Many authors will also send a copy if you email them.

Are preprints safe to cite?

Preprints have not yet passed peer review, so cite them with care and say they are preprints. Check whether a peer-reviewed version has since been published, and cite that version when it exists. Rules differ by field and journal.

What is the difference between Scopus and Web of Science?

Both are curated, multidisciplinary citation databases licensed by most universities. They differ in which journals they include and how they count citations, so results and citation counts will not match exactly. Many systematic reviews search both.

Do AI search engines replace database searching?

No. Tools such as Consensus or Semantic Scholar help you find starting papers quickly, but they cannot give you the documented, repeatable search that a methods section or a systematic review needs. Use them alongside a structured database search.

Related categories