Remove Accents (Diacritics)

About the Remove Accents (Diacritics)

Paste text and get it back without accents: crème brûlée becomes creme brulee, Łódź becomes Lodz, Nguyễn becomes Nguyen. Each character is split into its base letter and its accents with Unicode’s canonical decomposition, and the accents are dropped. Letters that are not a base letter plus an accent — ß, æ, œ, ø, ł, đ, þ and others — come from a table (ß → ss, æ → ae, ø → o, ł → l, þ → th). A table lists every change and how often it was made, so you can check the result.

Two more modes go further. Passport style follows the transliteration tables of ICAO Doc 9303, the standard for machine-readable passports: Müller → Mueller, Åse → Aase, Москва → Moskva, with the table’s rules for Ukrainian, Belarusian, Bulgarian, Macedonian and Serbian. Everything to ASCII also transcribes Greek and Cyrillic and spells punctuation and symbols in plain ASCII (“ ” → ", — → --, € → EUR, ½ → 1/2), for systems that accept nothing else.

Common uses

• Making file names, user names, URLs and database keys that work everywhere.
• Searching and matching names typed with and without accents.
• Filling forms, bank and booking systems that reject accented letters.
• Checking how a name will appear in a passport’s machine-readable zone or on a ticket.
• Removing Arabic tashkeel or Hebrew niqqud before searching or comparing text.

How accents are removed

Every character is decomposed with Unicode’s canonical decomposition (NFD, UAX #15): é becomes e plus a combining acute accent, ș becomes s plus a combining comma, ơ becomes o plus a horn. The accents are removed from Latin letters only, and the result is recomposed. Letters of other scripts keep their marks unless you ask: in Russian й and и are different letters, and in Hindi the marks are part of the spelling. Accents typed as separate marks are handled the same way as precomposed letters. Emoji are never changed.

Passport style: ICAO Doc 9303

ICAO Doc 9303 Part 3 (8th edition, 2021), section 6, has the transliterations “recommended for use by States” in the machine-readable zone of passports. Table A covers Latin letters: most simply lose their accents, while Ä → AE (or A), Å → AA (or A), Ö → OE (or O), Ü → UE (or UXX or U), Ñ → N (or NXX), Ø → OE, Æ → AE, Œ → OE, Þ → TH, IJ → IJ and ẞ (ß) → SS. Table B covers Cyrillic, for example Ж → ZH, Х → KH, Ц → TS, Ч → CH, Щ → SHCH, Ю → IU, Я → IA; the soft sign Ь is left out. Its notes change some letters by language: Ukrainian and Belarusian Г → H, Ukrainian И → Y and Є, Ї, Й, Ю, Я → YE, YI, Y, YU, YA at the start of a name, Belarusian Ё → IO, Bulgarian Щ → SHT, Macedonian Ѓ, Ќ, Џ → GJ, KJ, DJ, and Serbian Ж, Ч, Ш → Z, C, S. Serbian Ђ is D and Ћ is C.

The tables are recommendations: each country decides how it writes names in its passports, and many use their own national rules. The machine-readable zone itself is in capital letters, with < instead of spaces. (Doc 9303 Part 3 checked.)

Everything to ASCII

Accents are removed as above; Cyrillic follows ICAO Table B and Greek the transcription of the Greek standard ELOT 743, which the United Nations adopted in 1987: Αθήνα → Athina, Ευρώπη → Evropi, αυτό → afto, μπ → b at the start or end of a word and mp inside it (UNGEGN romanization report for Greek). Curly quotes become straight quotes, dashes become - or --, the ellipsis becomes ..., arrows become -> and <-, © ® ™ become (C) (R) TM, and currency signs become their ISO codes (€ → EUR, £ → GBP, ₹ → INR). Ligatures, fullwidth letters, superscripts and fractions are replaced by their plain forms (fi → fi, Ａ → A, ² → 2, ½ → 1/2), and the micro sign by u (5 µm → 5 um). Non-breaking and other special spaces become normal spaces, and invisible characters such as zero-width spaces are removed. Characters that have no ASCII spelling here — Chinese, Japanese, Hindi, Arabic, emoji — are replaced with ?, removed or kept, as you choose, and listed.

Limitations

• Works on up to 10 million characters at a time; split longer files into parts with the Text Splitter first.
• Removing accents can merge different words (Spanish papá, dad, and papa, potato); keep the original where the difference matters.
• Passport style follows ICAO’s recommended tables, not the rules of every country; check the passport itself when a booking must match it exactly.
• ICAO Table B lists Serbian under Г = H as well; Serbian Г is G in the Serbian Latin alphabet, so this tool keeps G.
• Chinese, Japanese, Korean, Indic, Arabic and Hebrew letters are not transliterated; a romanisation needs language rules (pinyin, Hepburn …) that this tool does not include.
• The ¥ sign is written Y, because it stands for both the Japanese yen and the Chinese yuan.

Privacy

Everything happens in your browser. What you enter or open here is not uploaded or stored by MySmartCoPilot.