Check any paper's citation impact

Field-weighted citation impact (FWCI) compares the citations a paper received in its publication year and the three years after it with the citations expected for comparable papers: works of the same type, published in the same year, in the same subfield. OpenAlex defines it as citations received divided by citations expected, so 1.0 is the world average, 2.0 means cited twice as often as expected, and 0.5 means half as often. Because of that four-year window, an older paper's FWCI describes how it was received when new, not what has happened since.

Because it compares a paper only with its peers, FWCI lets you set a pathology paper beside a pure-mathematics paper, which raw citation counts cannot: fields cite at very different rates.

Anything above 1.0 is cited more than the world average for comparable papers. Citations are very unevenly spread, with a small share of papers collecting most of them, so most papers sit below 1.0, and the citation percentile beside the FWCI is the clearer guide to where a paper ranks: a paper in the top 10% is cited more than at least nine in ten comparable papers.

For a paper published in the last three years, FWCI rests on a window that is still open and on few citations, so it can change a lot as new ones arrive: read it as provisional.

Paste the paper's DOI, a link to it, or its title into the box at the top of this page. The check shows the FWCI OpenAlex computes for that paper, with its citation percentile, its citations in five open indexes and its retraction record, with no account. OpenAlex has an FWCI for most papers but not all: some work types never get one.

Scopus and SciVal show their own FWCI to subscribers: in Scopus, open the paper's record and look beside its citation count.

Both follow the same recipe, citations in the publication year and the three years after it against the average for works of the same type, year and field, but they are computed on different data. OpenAlex counts citations across its open catalog of scholarly works and groups papers by its own subfields. Scopus counts only citations from the documents Scopus indexes and uses its own subject classification.

When a report or an application asks for the Scopus figure, use Scopus.

Each index counts citations only from the documents it holds, and each matches references to papers its own way. Google Scholar crawls publisher, university and repository sites, so it also counts theses, books, preprints and unpublished papers (and sometimes duplicates), and usually reports the highest number. Scopus and Web of Science count citations from the journals and books they select.

The five indexes here are open: OpenAlex and Semantic Scholar keep the broadest catalogs, Crossref counts citations from the reference lists publishers deposit with it, OpenCitations publishes open citation data, and Europe PMC covers biomedicine and the life sciences. They are not fully independent (OpenCitations and OpenAlex both draw on Crossref's reference lists), but the spread across them is a useful check on any single count.

The check reads OpenAlex's retraction flag and the retraction, correction and expression-of-concern notices Crossref links to the paper, and says which, if any, are on record.

A notice that neither record holds is invisible to it, so before relying on a result for a decision that matters, also read the article's page on the publisher's site.

No. A journal impact factor describes a journal: roughly, the average number of citations its articles from the previous two years received in a year. It says little about any one paper, because citations within a journal are very unevenly spread.

To judge a single paper, look at its own citations against comparable papers, which is what FWCI and the citation percentile measure.