Skip to content
100% local

Homoglyph detector

Find look-alike letters from other alphabets hiding in your text and fix or list them.

Input

Homoglyph detector

Paste any text and this tool scans it character by character for letters that look Latin but actually come from another alphabet — a Cyrillic а standing in for "a", a Greek ο standing in for "o", and dozens of similar pairs across the Cyrillic and Greek scripts. These substitutions are the core trick behind homograph attacks: a domain like "pаypal.com" or a username like "аdmin" can look identical to the real thing while pointing somewhere else entirely, and a glance at the screen will not catch it.

Switch the view between a findings list and corrected text. The findings list reports every match with its line number, the character itself, its Unicode code point and the script it belongs to, so you see exactly what was swapped and where. Corrected text instead returns your input with every flagged character replaced by its Latin counterpart. "Only mixed-script lines" narrows either mode to lines that combine Latin letters with look-alikes, so a paragraph written entirely in Cyrillic or Greek is not flagged just for using its own alphabet. "Ignore single-script text" goes further and skips the whole input when every letter already belongs to one script, leaving genuinely foreign-language text alone.

The detector recognises common Cyrillic and Greek look-alikes for Latin letters — the pairs actually used in spoofing, not the full Unicode confusables table, which runs into the thousands and would bury real findings in noise. Detection updates as you type, and even long documents finish in milliseconds.

Everything runs locally in your browser; the text you paste is never uploaded anywhere. That makes it safe to check sensitive material — internal usernames, domain lists, code snippets — before it ships. Copy the result, download it as a .txt file, clear the input, or send the output back into the input to run another pass.

FAQ

What counts as a homoglyph here?
A letter from another alphabet — currently Cyrillic or Greek — that is visually indistinguishable, or very close, to a Latin letter, such as Cyrillic а for "a" or Greek ο for "o". These are the pairs used in real domain- and username-spoofing attacks.
What does "corrected text" actually change?
It replaces every flagged character with its Latin counterpart and leaves everything else — spacing, punctuation, other letters — untouched. Use "List findings" instead if you want to review each match before anything is changed.
Why would a line in Cyrillic or Greek not get flagged?
Turn on "Only mixed-script lines" or "Ignore single-script text" and lines or whole texts that use just one alphabet are left alone. Ordinary text written in Russian, Ukrainian or Greek is not an attack — the tool only cares about scripts mixed together on the same line.
Does it catch every possible confusable character?
No. It covers the common Cyrillic and Greek look-alikes for Latin letters, the ones actually seen in spoofing, rather than the full Unicode confusables table, which runs into the thousands of mostly irrelevant entries.
Is my text uploaded anywhere?
No. Detection runs entirely in your browser — your text is never sent to a server.