Find and Replace Text Online - Regex, Whole Word, Match Case
Find and replace
How find and replace works
Paste any text — a document, a list, code or a CSV export — then type what to find and what to put in its place. The match count and the highlighted preview update as you type, and the result is ready to copy. Nothing leaves your browser.
How to use it
- 1 Paste or type your text in the Text box.
- 2 Enter the text to find and its replacement. Leave Replace with empty to delete every match.
- 3 Tick Match case, Whole word or Regular expression if needed, check the highlighted matches, then copy the result.
Whole word in every language
Whole word stops “cat” from matching inside “concatenate”. Many tools rely on the regex \b boundary, which only knows the letters A–Z: it treats “é” or “ş” as a word break, so “caf” is found inside “café”, and in Cyrillic or Hindi text it often finds nothing at all. This tool instead checks that there is no letter, accent mark, digit or underscore on either side, so it behaves the same in Turkish, Russian, Hindi or Vietnamese. Chinese and Japanese don’t put spaces between words, so there a whole-word match needs a space or punctuation around it.
Plain text mode: dots and dollar signs stay literal
With Regular expression off, everything you type is taken literally: searching for “3.5” finds only 3.5, not 305 or 3x5, and “(draft)” finds the parentheses themselves. The replacement is literal too, so “$10” is inserted as $10 instead of being read as a reference to group 1 followed by a zero.
Regular expressions and capture groups
In regex mode the pattern uses JavaScript syntax. Parentheses capture parts you can reuse: find (\d{4})-(\d{2})-(\d{2}) and replace with $3.$2.$1 to turn 2026-03-14 into 14.03.2026. Named groups work as (?<year>\d{4}) with $<year>, and $& inserts the whole match. ^ and $ mark the start and end of each line, and \n in the replacement inserts a line break. An invalid pattern shows an error instead of a wrong result.
Match case and special letters
Match case is off by default, so “Apple” and “apple” are both found. Case-insensitive search follows Unicode rules: “é” finds “É” and “я” finds “Я”, but the Turkish dotted “İ” and dotless “ı” are separate letters that “i” and “I” do not cover, so search for them explicitly. In regex mode an emoji counts as one character, so a single dot matches the whole emoji.