Конвертер омоглифов — обнаружение и очистка Unicode
Conversion is best-effort: mapped confusables and NFKC folding are deterministic, but some legitimate Unicode will not be flagged.
Paste or type — results update as you type (lightly debounced for long input).
Suspicious characters in the original view are underlined and labeled “susp.” in addition to highlight color.
No characters to analyze yet. Paste text above to see a homoglyph and confusables breakdown.
Обнаружение визуально похожих символов Unicode и преобразование в более безопасный ASCII или нормализованный Unicode. Бесплатно в браузере.
What this homoglyph converter does
A homoglyph converter detects Unicode characters that can look like ordinary Latin letters or digits (confusables) and shows how they were normalized or replaced. You can also run Generate homoglyphs to build deterministic spoof-like strings from plain ASCII for testing. This free confusables converter and unicode confusable checker runs entirely in your browser—use it as a unicode normalization tool, spoof text cleaner, or lookalike character converter for security and moderation.
How to use this unicode normalization tool
- Paste or type the text you want to inspect in the input area.
- Choose Preserve Readable Unicode, Strict ASCII Fallback, or Generate homoglyphs depending on whether you are cleaning pasted text or building test samples from ASCII.
- Compare the original (with highlights) and output side by side; review the analysis table for code points and reasons.
- Use Copy output text to copy the result, or Clear to reset.
Conversion modes: readable Unicode vs unicode to ASCII vs generate
- Preserve Readable Unicode applies the bundled confusable map first, then NFKC normalization when it changes a character. Output may still contain non-ASCII letters that were not flagged.
- Strict ASCII Fallback uses the same detection rules but prefers ASCII replacements from the map. Characters not in the map stay as-is unless NFKC changes them; this mode is stricter and better for spoof text cleaner workflows.
- Generate homoglyphs walks your input left to right and replaces each ASCII letter or digit with a single deterministic homoglyph from the same bundled map (inverse of cleanup). Characters without a chosen substitute stay ASCII.
What counts as suspicious here
In cleanup modes, a character is flagged when it is a known confusable with a defined replacement, or when NFKC normalization alters it (for example fullwidth digits). In generate mode, “suspicious” marks each position where an ASCII letter or digit was replaced. This is not a full unicode security audit—only deterministic rules.
Common use cases
- Checking spoofed usernames, domains, or homoglyph attacks in pasted strings.
- Moderation and trust-and-safety review of lookalike character tricks.
- Developer debugging of encoding issues alongside a homoglyph detector view.
Limitations and safety notes
The bundled map covers common Cyrillic, Greek, and fullwidth lookalikes—not every Unicode confusable. Generate mode is for authorized testing only; do not use it to deceive people or bypass protections. Results are deterministic, not ML-based risk scores. Always combine with human review for high-stakes decisions.
Privacy
All detection and conversion run locally in JavaScript after the page loads. No text is sent to a server for processing.
Часто задаваемые вопросы
What is a homoglyph?
A homoglyph is a character that looks like another character from a different script or encoding, such as a Cyrillic “а” that resembles Latin “a”. Attackers use them in spoofed domains, usernames, and messages. This tool surfaces those characters with code points and replacements.
How is this different from Unicode normalization?
Unicode normalization (for example NFKC) folds compatibility and width variants in a standard way. This tool combines NFKC with a small explicit confusable map so you see why each character was flagged—useful beyond a plain unicode normalization tool pass.
Will strict ASCII mode change legitimate non-English text?
Strict ASCII Fallback replaces characters that appear in the bundled confusable map or change under NFKC. Other non-ASCII letters are left unchanged. Legitimate text can still change if it contains those mapped characters, so review the analysis list.
Is my text sent to a server?
No. After the page loads, detection and conversion use JavaScript only. This confusables converter does not upload your paste for analysis.
Can this catch every spoof?
No. Coverage is deterministic and map-based, not machine learning. It helps with common lookalike character tricks but is not a full security audit.
What does Generate homoglyphs mode do?
It replaces ASCII letters and digits in your input with deterministic lookalikes from the same bundled map used for cleanup—useful for building test strings in authorized environments (for example moderation QA). It is not for impersonation or bypassing protections.