Text Difference Checker
Compare two texts character by character and reveal hidden Unicode, spacing, punctuation and normalization differences.
Your text never leaves your browser.
Add text to both panels to compare them character by character. Hidden spaces, look-alike letters, dash and quote variants and Unicode normalization differences will be highlighted here.
Why do identical-looking texts not match?
Computers compare text as a sequence of Unicode code points, not as the shapes you see on screen. Unicode has dozens of space characters, several kinds of dashes and quotes, letters from different alphabets that look the same, and more than one way to encode accented letters.
When text passes through PDFs, web pages, spreadsheets, chat apps or word processors, any of these can be swapped in silently. The result looks identical but fails equality checks, lookups and searches. This tool shows exactly which characters differ and why. To find and clean hidden characters in a single text, use the Invisible Character Detector.
Common hidden differences
| Text A | Text B | What changed |
|---|---|---|
| spaceU+0020 | NBSPU+00A0 | Non-breaking space from web pages and word processors |
| nothing— | ZWSPU+200B | Zero-width space hidden inside a word or number |
| -U+002D | –U+2013 | Hyphen versus en dash, often auto-corrected |
| "U+0022 | “U+201C | Straight versus curly quotation mark |
| aU+0061 | аU+0430 | Latin a versus Cyrillic а — a look-alike |
| éU+00E9 | éU+0065 U+0301 | Same letter, precomposed versus decomposed (NFC vs NFD) |
| LFU+000A | CR LFU+000D U+000A | Unix versus Windows line endings |
What is Unicode normalization?
Many accented letters can be stored two ways: as a single precomposed character, or as a base letter followed by a combining mark. Both render identically, but their code points differ, so a plain comparison says they are not equal.
NFC (composed) combines letters and marks wherever possible and is the usual choice for storing and comparing text. NFD (decomposed) splits them apart. When the only differences are normalization differences, CleanGlyph tells you the texts are equivalent and offers to normalize them to NFC.
When to use a text difference checker
Excel & Google Sheets
Find why VLOOKUP, XLOOKUP or MATCH says two cells are different when they look identical.
Usernames & emails
Spot look-alike letters and hidden characters that create duplicate or spoofed accounts.
Config values & environment variables
Compare configuration values, environment-variable contents, IDs and reference strings for hidden spaces, punctuation or Unicode differences.
Database values
Explain failed equality checks, duplicate rows and unique constraint surprises.
Frequently asked questions
Why do two identical-looking strings not match?
Software compares Unicode code points, not how text looks. A non-breaking space instead of a normal space, a zero-width space, a curly quote or a letter from another alphabet makes two strings unequal even when they look the same on screen.
How do I compare two strings character by character?
Paste one value into Text A and the other into Text B. CleanGlyph compares every Unicode code point and highlights each difference, with the character names, code points and exact positions on both sides.
Can invisible characters make strings different?
Yes. Zero-width spaces, zero-width joiners, soft hyphens and byte order marks take up no visible space but are still characters. One string can contain them while the other does not, so equality checks, lookups and searches fail.
What is Unicode normalization?
Some characters can be written in more than one way. For example, é can be one precomposed character (U+00E9) or the letter e followed by a combining accent (U+0065 U+0301). Normalization forms such as NFC and NFD convert text to one consistent encoding so equivalent strings compare as equal.
Why does Excel say two values are different?
Usually one cell contains a hidden character, trailing space, non-breaking space or a different dash or quote. Paste both cell values here to see exactly which character differs, then clean or normalize the source data.
Can different Unicode characters look the same?
Yes. Latin a and Cyrillic а, Latin o and Greek ο, and several dash and space characters are visually almost identical. These look-alikes are sometimes used in phishing or spoofed usernames, which is why CleanGlyph marks them for review.
Does this tool upload my text?
No. The comparison runs entirely in your browser using JavaScript. Your text is never sent to a server, stored or logged.