Search the indexed article corpus by title. Every result carries its field-normalized citation impact — so a mathematics paper and an immunology paper can be compared honestly.
Enter a search term to begin.
Covers 101.7 million articles and reviews from 59,872 of the 62,643 journals in the catalogue — 1,276 fewer journals than before the corpus was rebuilt on the 23 September 2026 OpenAlex release, which took 2.9 million articles out of it. This said “7.2 million — the 25 most-cited of every indexed journal, plus the most-cited works of every author” until 19 September 2026, and that is what the corpus was: a selection. The article layer moved to our own server that day and stopped being one. It has not gone back to being one: re-measured on the served table 27 September 2026, 100,287,299 rows still sit beyond their journal’s 25th, and the deepest journal now runs to 1,781,398 ranked rows — deeper than before, in a corpus that got smaller. Those two read 103,183,183 and 1,031,216 on 19 September, when the corpus held 104,664,791. It is still not the entire literature: where a journal is outside the catalogue, its articles are outside this corpus too. Retracted work is outside this corpus by design, so a retracted paper will not be found here and its absence is not a verdict on it. This said “on both sides of every citation” until 30 Aug 2026, and then said the reverse: that only the cited side was filtered, so citations made BY a retracted paper still counted toward every CNI and percentile here. That was true of the edition being served when it was written, and it has been false since 3 September 2026, when v1.1-rc1 was promoted and extended the retraction filter to the citing side as well. v1.1-rc1 is still the edition behind every CNI, percentile and quartile on this site — measured 28 September 2026, cl_quartiles serves v1.1-rc1 — so both sides are filtered today and this sentence was wrong for twenty-five days. Errata E64; the methodology page’s edition box has carried the correction since the promotion and this sentence did not follow it. — see the methodology. Source: OpenAlex (CC0). CNI normalizes citations against each article's own subfield, year and document type (1.00 = world average); Pct is its percentile in that peer group. Per DORA, these describe individual articles, not their authors.