Remove Diacritics
Take the harakat out of Arabic and Urdu, the niqqud out of Hebrew, and the accents out of Latin, Greek and Russian text. Choose which marks go; letters that need their mark stay. Everything happens on your device.
How to remove diacritics
- Type or paste your text. It can mix scripts: Arabic, Urdu, Hebrew and Latin are each handled by their own rules.
- The text without diacritics appears below as you type, with a count of what was removed.
- Untick any kind of mark you want to keep, such as shadda or the Hebrew shin dot, then copy or download the result.
Examples
- اَلسَّلَامُ عَلَيْكُمْbecomesالسلام عليكم
- پاکِسْتان زِنْدَہ بادbecomesپاکستان زندہ باد
- שָׁלוֹם עֲלֵיכֶםbecomesשלום עליכם
- Ça va? Où est le café?becomesCa va? Ou est le cafe?
- Łódź, Kraków, GdańskbecomesLodz, Krakow, Gdansk
- Где моя́ ру́чка?becomesГде моя ручка?
How we checked it
- Latin: on 17,600 words with accents from 47 languages, plain ASCII matches ICU's Latin-ASCII transform except on a few IPA letters such as ɑ and ʃ, which ICU leaves as they are; German spelling matches ICU's de-ASCII on every German word.
- Arabic script, Hebrew and Russian: we removed the marks from words as Wiktionary writes them with their vowels or stress and compared the result with the entry's plain title. It matched on 4,975 of 5,000 Arabic words, 4,968 of 5,000 Urdu, 3,942 of 3,999 Russian and 681 of 731 Persian. Almost all the rest differ only in spacing, punctuation or letter forms (ك for ک, a missing ZWNJ), or the title is another form of the word.
- Hebrew matched on 4,047 of 5,000 words. Nearly all the others are spelled in full without niqqud, with an extra ו or י for the vowels (אֱלֹהִים is written אלוהים). Adding those letters takes knowing the word, so the tool leaves the spelling as it is.
Frequently asked questions
Which Arabic and Urdu marks does it remove?
The short vowels (zabar, zer and pesh, called fatha, kasra and damma in Arabic), tanween, jazm (sukun), shadda, the superscript alef (khari zabar), Quranic signs and tatweel, each with its own box. It keeps the hamza and madda that are part of a letter, so آ, أ, ئ, ؤ, ۂ and ۓ stay as they are: آئینہ and سؤال. Tick "Hamza and madda on alef" to write أ, إ and آ as a plain ا, as some search indexes do.
Why does ٱلْكِتَٰبُ become الكتب and not الكتاب?
In the Quran's spelling the long a is often only the small superscript alef, so removing it leaves الكتب. Modern spelling writes it as a full alef. For Quranic text, tick "write ٰ as ا": الكتاب. Words that modern spelling also writes without the alef keep it out: هذا ذلك.
Why doesn't Hebrew come out in the usual spelling?
Removing niqqud gives the spelling without vowel letters (ktiv haser): אלהים. Everyday Hebrew without niqqud adds ו and י as vowel letters (ktiv male, אלוהים). Which ones to add depends on the word, so the remover doesn't guess: of 5,000 Wiktionary words we checked, 17% differ from their usual spelling in this way.
Can I keep ñ, or write ä as ae?
Yes. Type the letters to keep in "Keep these letters", for example ñÑ for Spanish: mañana stays mañana. For German, choose the German spelling and Müller becomes Mueller; for Danish and Norwegian, Ålborg becomes Aalborg. "Plain ASCII letters" also spells ß, æ and þ in plain letters: Strasse.
Does it remove ł, ø and đ?
Yes: Lodz, Tromso, Dorde. These letters have no separate mark in Unicode, so many removers that rely on splitting letters from marks leave them as they are.
Will it break Hindi, Bangla or other scripts?
No. Only Arabic-script, Hebrew, Latin, Greek and Cyrillic marks are removed. The vowel signs of Hindi, Bangla, Thai and other scripts are part of their letters, so text in those scripts comes back exactly as it was.
Is my text uploaded?
No. Everything runs in your browser, and your text never leaves your device. The page works offline once it has loaded.