How do I use capture groups in the replacement?

Tick Regex, then reference groups with $1, $2 and so on. Searching for (\w+), (\w+) and replacing with $2 $1 flips "Smith, John" into "John Smith" on every line at once. Named groups work too: (? \d{4}) is referenced as $ . With Regex off the replacement is literal, so a dollar sign stays a dollar sign.

Why does whole-word matching fail on Cyrillic in other tools?

Most of them wrap your search in \b, and in JavaScript \b only knows the ASCII letters A–Z. Search for кот inside котёнок with a \b-based tool and it matches, because ё is not an ASCII word character. This tool checks the surrounding characters against the full Unicode letter and digit categories instead, so Cyrillic, Greek and accented Latin words bound the way you would expect.

What does Match case of original do that Case sensitive does not?

They are different jobs. Case sensitive decides what gets found: with it off, Color and color are both hits. Match case of original decides what gets written: replacing color with colour turns COLOR into COLOUR and Color into Colour instead of flattening everything to lowercase. Turn both on and only exact-case matches are touched.

Can I replace line breaks or tabs?

Yes, with Regex on. Use \n for a newline, \r\n for Windows line endings, \t for a tab and \s+ for any run of whitespace. Replacing \n{2,} with a single space collapses blank-line gaps. Anchors work per line, so ^ matches the start of every line rather than only the start of the text.

Is there a size limit?

The tool stops after 100,000 matches in one pass and says so rather than silently truncating. Text size is limited only by your device's memory — a few megabytes of log file is fine. A 500 KB file with a couple of thousand hits finishes in a few milliseconds.

Does my text get uploaded anywhere?

No. The whole search and replace runs in JavaScript inside your browser tab. There is no upload step, no server queue and no logging. Load the page, disconnect from the internet, and it keeps working — which matters when the thing you are cleaning up is a database dump or a client contract.

Find and Replace Text Online — Regex, Whole Word, Case

Find

Replace with

Input

Result

Matches found

Replacements made

Lines changed

Paste text, type what to find and what to put in its place, and the result updates as you type. Regular expressions with capture groups, whole-word matching that understands Cyrillic, and an option to keep the original capitalisation are one checkbox each. Nothing is uploaded — the replacement happens inside your browser tab.

How it works

1
Paste your text Drop anything into the input box — a paragraph, a CSV export, a log file, a chunk of source code. Line breaks and indentation are preserved exactly.
2
Type the search and the replacement Fill in Find and Replace with. The result panel and the match counter update on every keystroke, so you can see how many hits a pattern gets before committing to it.
3
Turn on the options you need Regex unlocks capture groups and $1 references. Whole word stops cat from matching inside category. Match case of original keeps ALL CAPS and Capitalised hits capitalised after the swap.
4
Copy the result Hit Copy to put the output on your clipboard, or Use as input to feed it straight back in and run a second replacement on top of the first.

Your data stays private

All processing happens entirely in your browser. No files, text, or data are ever sent to our servers. You can disconnect from the internet and this tool will still work.

Frequently asked questions

How do I use capture groups in the replacement?: Tick Regex, then reference groups with $1, $2 and so on. Searching for (\w+), (\w+) and replacing with $2 $1 flips "Smith, John" into "John Smith" on every line at once. Named groups work too: (?<year>\d{4}) is referenced as $<year>. With Regex off the replacement is literal, so a dollar sign stays a dollar sign.
Why does whole-word matching fail on Cyrillic in other tools?: Most of them wrap your search in \b, and in JavaScript \b only knows the ASCII letters A–Z. Search for кот inside котёнок with a \b-based tool and it matches, because ё is not an ASCII word character. This tool checks the surrounding characters against the full Unicode letter and digit categories instead, so Cyrillic, Greek and accented Latin words bound the way you would expect.
What does Match case of original do that Case sensitive does not?: They are different jobs. Case sensitive decides what gets found: with it off, Color and color are both hits. Match case of original decides what gets written: replacing color with colour turns COLOR into COLOUR and Color into Colour instead of flattening everything to lowercase. Turn both on and only exact-case matches are touched.
Can I replace line breaks or tabs?: Yes, with Regex on. Use \n for a newline, \r\n for Windows line endings, \t for a tab and \s+ for any run of whitespace. Replacing \n{2,} with a single space collapses blank-line gaps. Anchors work per line, so ^ matches the start of every line rather than only the start of the text.
Is there a size limit?: The tool stops after 100,000 matches in one pass and says so rather than silently truncating. Text size is limited only by your device's memory — a few megabytes of log file is fine. A 500 KB file with a couple of thousand hits finishes in a few milliseconds.
Does my text get uploaded anywhere?: No. The whole search and replace runs in JavaScript inside your browser tab. There is no upload step, no server queue and no logging. Load the page, disconnect from the internet, and it keeps working — which matters when the thing you are cleaning up is a database dump or a client contract.

From the blog

Find and Replace Text Online: Regex, Whole Words, and Case How to run a bulk find and replace in the browser — capture groups, the word-boundary trap that breaks non-English text, and keeping capitalisation intact. Read the post →

Find and Replace Text Online — Regex, Whole Word, Case

How it works

Your data stays private

Frequently asked questions

From the blog

Related tools