Help & FAQ

OpenAlex, Semantic Scholar, PubMed, Europe PMC, Crossref, arXiv and DOAJ on every search, plus specialist databases when your topic calls for them: DBLP for computer science, INSPIRE-HEP for physics and ERIC for education. Results found in several databases are merged into one card.

Yes — that's what it's designed for. It recognises hundreds of field-specific terms and acronyms (GWAS, DFT, QCD, fMRI…) and uses PubMed's MeSH vocabulary for medical terms. It works out which words in your question are the most specific (by checking how rarely they appear across 250 million papers) and requires strong results to contain them. It also detects the research area your results belong to and searches inside it.

When only a few papers match every specific part of your question, Kanora automatically follows the citation network of the best matches — papers they cite, papers that cite them, and closely related work — and retries without the least specific word. You can also start one yourself with “Run a deep search”. Papers found this way are labelled “Via citation network”.

Be specific about who, what and where: “screen time and sleep in adolescents” beats “screens”. Put exact phrases in quotes ("working memory") and exclude words with a minus sign (-mice). Kanora ranks papers that match every part of your question first, and hides papers that only match a small part of it.

Researchers often use different terms from everyday language — “adolescents” rather than “teens”. Kanora knows many of these equivalents, searches for both, and tells you which scholarly terms it added.

No, and we don't pretend otherwise. Preprints (for example from arXiv) are labelled “Preprint · not yet peer reviewed”. Use the “Hide preprints” filter if you only want published work, and always check the journal yourself.

The databases didn't report a free version. Try your library's login, the “Scholar” button (which often finds free copies), or the author's own website.

From the titles, metadata and abstracts of your top results. Findings and quotes are copied word-for-word from abstracts and linked to their source. Methods are detected from the abstract's wording, so check the full paper before relying on them.

Yes — during early access every feature is free and there's no checkout. Paid plans will be announced before they launch.

No. Your saved articles, searches, folders and drafts are stored in your own browser. Use “Export backup” on the Dashboard to keep a copy or move it to another device.

APA 7, MLA 9, Chicago (notes-bibliography) and BibTeX. Click a style on any article card to copy it.

They're built from database metadata, which is sometimes incomplete (missing page numbers or unusual author formats). Always check citations against the original paper and your instructor's requirements.

In your browser's local storage on this device. Clearing your browser data deletes it, so export a backup from the Dashboard regularly.

No. The Article Analyzer and Study Tools read PDFs inside your browser. Your files and essays are never uploaded.

No — by design. Kanora helps you find sources, understand evidence and plan your argument, but the thinking and writing stay yours. Check your institution's rules on using study tools.