Bulk accepts 5,000 lines including blank lines, trims row edges, skips blank lines and preserves order with LF output. Results appear only when the whole batch succeeds. Turn bulk off for the original full output. Slugs treat punctuation and whitespace as separators and preserve Unicode by default; optionally strip Latin accents or remove apostrophes. A nonblank row with no retained letters/numbers fails the batch with its original line number, unlike the source which skips empty slugs. CompareTwoTexts Modes checked: 2026-10-28. Unicode casing / abbreviation handling is heuristic, not transliteration.
Stopword list
a, an, and, as, at, be, by, for, from, in, is, it, of, on, or, that, the, this, to, was, with
Each field: 1 MiB UTF-8 (SQL: 100,000 characters); output: 2 MiB. Loading stops after 15 seconds; processing after 2 seconds. Local processing, without uploads or storage.
Default preserves Unicode with NFC. Optional NFD removes U+0300–036F accents after Latin letters, without transliterating other scripts. Length counts graphemes, capped at 10,000, trimming trailing separators. AP / Chicago 18 use finite English word lists, capitalizing first/last and subtitle openings. Four-letter prepositions capitalize in AP but lowercase in Chicago; five or more capitalize in both. Review phrasal verbs, brands and compounds manually; this is not a grammar parser. Chicago / AP
Enter input or load a sample, then run.