Written and maintained by CASRAI Editorial Board
Last updated
Keenious is an academic search engine that recommends relevant research papers based on your own writing, using AI. You use it in a Word or Google Docs sidebar, or by pasting a document into keenious.com. This guide explains what it searches, what it does not, the exact query syntax its help center documents, and how to disclose its use. It is written for researchers and students deciding where a recommender like this fits in a literature search.
What Keenious is
Per its help center, Keenious reads the text you are working on and returns papers that match its meaning, rather than requiring you to build a keyword query first. It is offered as a Word add-in (Office Add-in Store / AppSource), a Google Docs add-on, and a web version where you paste any document. Its help articles also cover bookmarks, exporting results and bookmarks, follow-up suggestions, searching across languages, and searching for authors, venues and institutions.
Who runs it and what it covers
The help pages used for this guide do not name an operator, so this page attributes everything to Keenious’s own help center (help.keenious.com) and does not go further.
Data source: OpenAlex
Keenious states that its data comes from OpenAlex, a free, open catalog of scholarly works run by the non-profit OurResearch. Keenious does not index everything in OpenAlex: it searches a curated subset. Its help pages give the size as roughly 180-190 million works (one article says around 181 million, another about 188 million publications), covering more than 100 languages.
Keenious describes four layers on top of the raw OpenAlex data:
- Curation: junk records, duplicates and records from predatory venues are removed.
- Peer-review labels: taken from the Norwegian Scientific Index.
- Field-weighted citation impact (FWCI): a field-normalised citation measure.
- Language detection: across 170+ languages.
What is actually searched
Keenious searches titles and abstracts, not full text. It combines meaning-based matching (embeddings of titles and abstracts) with keyword matching (BM25). A paper whose abstract is missing is therefore harder to surface.
How to search
- Open the sidebar in Word or Google Docs, or go to keenious.com and paste a document.
- Let it read your text, or select the passage you want matched.
- Choose a Scope: Focused, Balanced (the default), Exploratory or All.
- Apply filters if needed: publication year, document type, peer-review status, open access. Keenious says filters are hard constraints, so they remove results rather than re-rank them.
- Refine with syntax (below), or optionally pick a research area. Research areas are clusters named by a language model; selecting one filters the list without changing ranking.
- Bookmark useful papers and use the export function for results and bookmarks.
Search syntax documented by Keenious
| You type | Effect |
|---|---|
| “quoted phrase” | The phrase becomes a hard requirement |
| AND / OR (uppercase only) | Boolean combination of terms |
| -term or NOT term | Excludes the term |
| A bare four-digit year | Boosts papers from that year (not a hard filter) |
Ranking is relevance first. Citations, recency, FWCI, peer-review status, venue type and language match are described as modest adjustments. For general Boolean technique, see Boolean search operators for literature searching.
Strengths and limits
Where it helps
- Starting from a draft paragraph instead of guessing keywords, which suits the early, exploratory stage of reading around a topic.
- Searching without leaving the document you are writing in.
- Searches across 100+ languages, with language matching in the ranking.
Limits Keenious itself states
- It describes itself as intended for discovery rather than as a complete bibliographic record. For citation studies or systematic reviews, its help center points to using OpenAlex directly.
- Titles and abstracts only: no full-text search.
- Citation counts run lower for non-Western and non-English work, which can affect the modest citation-based adjustments.
- Some records are mislabeled, and records without abstracts are harder to find.
Because the index is a curated subset and Keenious describes itself as a discovery tool, it is better treated as a supplement than as the sole search in a protocol-driven review (editorial view). See using AI in a literature review without breaking the method for how to position such tools.
Documenting your use of Keenious
Keenious’s help center advises disclosing tool use and notes that requirements vary between fields and programs. It lists four places to do so: the methodology section, the body text, the literature-review approach, or a detailed description. Its suggested wording is along the lines of: “Searches were conducted in [databases], and Keenious was used to identify additional relevant articles.” Check your journal’s or institution’s policy before choosing the form.
Alternatives
The table groups alternatives by what they are, not by a quality ranking. Each category links to the site’s own explainers.
| Option | Category | Where to read more |
|---|---|---|
| OpenAlex | The open catalog Keenious is built on; queried directly with filters and an API | OpenAlex: what it is and how it works |
| Google Scholar, Semantic Scholar, CORE, BASE, Dimensions | Scholarly metadata search engines | Metadata search engines compared |
| Elicit, Consensus, SciSpace | AI research assistants | Elicit vs. Consensus, Consensus vs. SciSpace |
FAQ
What is Keenious?
An academic search engine that recommends research papers from the text you are writing, using AI, available as a Word add-in, a Google Docs add-on and a web app.
Where does Keenious get its papers?
From OpenAlex, filtered into a curated subset of roughly 180-190 million works according to its help pages.
Does Keenious search full text?
No. It searches titles and abstracts.
Can I use Keenious for a systematic review?
Keenious says it is for discovery, not a complete bibliographic record, and recommends OpenAlex directly for systematic reviews and citation studies. See systematic literature review tools for review-oriented software.
How do I search for an exact phrase or exclude a word?
Put the phrase in quotes for a hard requirement, and use -term or NOT term to exclude. AND and OR must be uppercase.
Do I have to say I used Keenious?
Requirements vary by field and program; Keenious advises disclosing, and provides example wording.
Sources
- Keenious Help Center, “What is Keenious” (help.keenious.com/en/articles/185112), read 2026-11-07.
- Keenious Help Center, “How search works in Keenious” (help.keenious.com/en/articles/169371), read 2026-11-07.
- Keenious Help Center, “Understanding our data source: how Keenious uses OpenAlex” (help.keenious.com/en/articles/203843), read 2026-11-07.
- Keenious Help Center, “Documenting the use of Keenious in your work” (help.keenious.com/en/articles/211334), read 2026-11-07.
