Unicode Confusable Character Detector

Detect Unicode lookalike characters online

This Unicode confusable detector inspects code points, identifies scripts, detects mixed substantive scripts, uses Unicode confusable mappings, generates a comparison skeleton, shows likely visual lookalikes, supports an optional trusted target, reports invisible and directional characters separately, and provides review indicators rather than intent judgments. Analysis is processed on the server.

Confusable character example

The default example раypal.com begins with Cyrillic U+0440 CYRILLIC SMALL LETTER ER and U+0430 CYRILLIC SMALL LETTER A. With the trusted comparison paypal.com, the exact strings differ, but the confusable skeletons match when using the bundled mapping data.

Plain ASCII paypal.com has no non-ASCII confusable findings and does not mix substantive scripts.

Mixed scripts vs Unicode confusables

Mixed scripts mean more than one writing system is present. Confusables are characters or sequences that may look alike. A mixed-script string may be legitimate, and a same-script string may still contain confusables. Script detection and confusable mapping are separate analyses. Common and Inherited characters, such as punctuation, digits and combining marks, do not automatically create a mixed-script warning.

How confusable skeletons work

A skeleton is a comparison representation generated from the bundled Unicode Technical Standard #39 confusables mapping data. Lookalike strings may reduce to the same skeleton, but skeleton equality is a review signal, not proof of malicious intent. The skeleton is not corrected text, and results depend on the Unicode data version and mapping algorithm used.

Cyrillic, Greek and Latin lookalikes

Common examples in the bundled data include Cyrillic р mapping to Latin p, Cyrillic а mapping to Latin a, and Greek Α mapping to Latin A. Fonts and rendering affect how close these look. Same-script mappings also exist; for example Latin ǃ maps to the common exclamation mark !, which shows why mixed-script detection alone is insufficient.

Invisible and directional characters

Some Unicode characters are invisible or affect text direction. Zero-width spaces, joiners, soft hyphens, bidirectional controls, directional isolates, variation selectors, combining marks and control characters are reported as separate indicators. They may be legitimate, but they can complicate code review, identifiers, domains and logs.

Confusable domains and Punycode

This tool analyzes Unicode characters. It does not verify domain ownership, check registration status, confirm phishing intent or fully emulate browser IDN display policies. Punycode conversion and IDNA validation are separate operations, and a Unicode string may have a Punycode representation. Use the Punycode Converter for IDNA/Punycode conversion.

Limits of automated spoofing detection

Visual appearance depends on font and rendering. Context matters, mixed scripts can be legitimate, and equal skeletons do not prove intent. The detector does not authenticate a sender, domain, file or account, and it does not replace browser, registry or application IDN policies.

How to use the detector

1. Paste the Unicode text or domain-like string to analyze.
2. Optionally enter a trusted comparison spelling.
3. Run the character analysis.
4. Review scripts, flagged code points, lookalike mappings and indicators.
5. Compare skeletons when a trusted target is supplied.
6. Copy or download the report for further review.

Privacy and server processing

The submitted text is sent to UnicodeNow for confusable-character analysis. Input is processed securely and is not intentionally retained after the request completes.

Do not submit passwords, private keys, access tokens, personal records or other sensitive data.

Frequently asked questions