What it does
This tool removes diacritical marks from Latin-script text. É becomes E, ñ becomes n, ü becomes u, and ç becomes c. It is the cleanup step you need when a name, address, or title has to fit into a system that only stores plain ASCII, or when you want a search key that matches both the accented and unaccented spelling.
It is not a translation or a language model. It does not know that a particular spelling is preferred, and it does not decide which of two names is correct. It applies a deterministic Unicode transformation, reports how many characters changed, and leaves everything it does not recognize alone.
How accents are removed
Unicode can represent é as a single precomposed character or as e followed by a combining acute accent. The tool normalizes the text to the decomposed form, removes the combining marks, and recomposes the result. That handles the common accents in one pass: acute, grave, circumflex, tilde, diaeresis, cedilla, caron, ring, and their relatives.
Because the operation is character-based, it works on any amount of text and does not need a dictionary. It also preserves line breaks, capitalization, and punctuation. The character count can get shorter when a combining sequence becomes one letter, but the visible words are the same.
Special letters
Some Latin letters are not a base letter plus a mark. Æ, Œ, Ø, Ł, Đ, and ß are distinct letters, so normalization alone leaves them in place. Turn on “expand special letters” to map the common ones to ASCII: æ to ae, œ to oe, ø to o, ł to l, đ to d, and ß to ss.
That mapping is a convenience, not a universal transliteration standard. Danish ø and Norwegian ø may be better represented differently in a specific catalog, and German ß is sometimes written ss and sometimes left alone. If a downstream system has its own rules, use them. The option gives you a sensible default and shows exactly what it changed.
Examples
Before
Café, naïve, résumé
Straße and Æsir
ŁódźAfter
Cafe, naive, resume
Strasse and AEsir
LodzCafé becomes Cafe, naïve becomes naive, and résumé becomes resume. Straße becomes Strasse because ß expands to ss, Æsir becomes AEsir, and Łódź becomes Lodz. The line breaks and capitalization are preserved. Without the expansion option, Straße, Æsir, and Łódź would keep their special letters.
Before
“Hello,” she said. “It’s fine.”After
"Hello," she said. "It's fine."Curly quotation marks and apostrophes are common in text copied from a word processor. Some code, CSV, and database workflows expect the straight ASCII characters. The optional normalization converts the curly single and double quotes but does not remove them, so the sentence still reads the same.
Limits worth knowing
- Unaccented is not always correct. A person’s name, a brand, and a place name may be officially spelled with the mark. Use the result as a key or a fallback, not as a replacement for the display name.
- It is Latin-focused. Greek, Cyrillic, Arabic, Hebrew, Devanagari, Chinese, Japanese, Korean, and other scripts are not transliterated. They pass through unchanged unless their own combining marks can be removed safely.
- It is not a security tool. Removing accents does not normalize all lookalike characters, and it does not protect against homoglyph attacks. Two strings that look alike after this conversion are not necessarily the same identifier.
- Some languages treat the mark as a separate letter. In Swedish, å and ä are distinct letters, not decorative accents. Stripping them changes the alphabet and can change the meaning. Keep the original when the language matters.
- No automatic filename safety. The tool removes accents, not slashes, colons, question marks, or other characters that a filesystem may reject. Use a dedicated slug or filename cleaner for the rest.
Page updated