Remove accents from text

Strip accents, cedillas, umlauts, tildes and other diacritical marks from a text: “café” becomes “cafe”, “Façade” becomes “Facade” and “Łódź” becomes “Lodz”. Everything else is kept as it is, from punctuation and capital letters to emoji. The result appears as you type and the text never leaves your browser.

Unicode describes an accented letter as a base letter followed by one or more combining marks: é decomposes into e and an acute accent. The tool applies this canonical decomposition (the Unicode NFD form) to every Latin, Greek or Cyrillic letter, removes the diacritical marks and recomposes the letter. An accent typed as a separate character, common in text copied from macOS or a PDF, is removed in the same way.

Some letters do not decompose because their mark is a stroke rather than an accent. ø, ł, đ, ħ and ŧ are therefore replaced explicitly with o, l, d, h and t. On the other hand ß, æ, œ, þ and ð are not accented letters but ligatures or letters in their own right; they stay unchanged unless you turn on the option that spells them with plain letters (ss, ae, oe, th, d). A capital followed by lowercase letters is then written in title case: Ærø becomes Aero.

Only diacritics are removed. Greek stays in the Greek alphabet (ά becomes α) and Cyrillic in the Cyrillic alphabet, losing only its stress marks. й, ё, ї and ў are letters distinct from и, е, і and у and are not changed. Arabic vowel marks, Hebrew points, the vowel signs of Indian scripts, Korean, Chinese and emoji are copied unchanged, byte for byte: removing them would alter the text rather than simplify it.

Removing accents is not a full transliteration to ASCII. Curly quotes, the curly apostrophe, dashes and symbols remain. For a file name or a web address, the slug generator also converts spaces and punctuation.