Felo

Unicode Lookalike Character Converter Online

Replace Latin letters with visually similar Unicode homoglyphs, or normalize lookalike text back to plain Latin. Every mapped character is listed explicitly, unmapped characters such as digits, punctuation, emoji and non-Latin scripts pass through untouched, and each run includes a per-character change report. Try it online.

What is Unicode Lookalike Character Converter Online?

It is a text tool that swaps ordinary Latin letters for Unicode characters that look almost the same, and can also run the reverse direction to turn lookalike text back into plain Latin letters. The visible result often reads normally while the underlying code points differ. Unlike a simple two-box page, this agent states its full mapping, preserves everything it does not map, and reports each change.

Key Features

Explicit letter map. All 26 lowercase Latin letters have a defined homoglyph, and uppercase coverage is stated exactly: only A maps (to 𐊠, U+102A0), while uppercase B–Z pass through unchanged. This is a deliberate design choice: the reference table left those uppercase entries blank, and this agent treats blank entries as unmapped passthrough rather than deletion. No letter is silently skipped or invented.
Two directions. Convert Latin to lookalikes, or normalize lookalike text back to Latin.
Preservation guarantee. Digits, punctuation, spaces, emoji and non-Latin scripts are returned unchanged.
Change report. Each converted character is listed with its original code point and replacement code point.
Detection mode. Ask the agent to flag lookalike characters already present in your text without changing anything.

How to Use

Paste your text and say what you want: "convert to lookalikes", "normalize to Latin", or "detect lookalikes". If you only paste text, the agent assumes conversion to lookalikes.
Read the copyable output block first; the change report follows underneath.
Ask for a different variant, a partial range, or a re-run on edited text in the same conversation.

Example Use Cases

Testing filter behavior. Convert a sample phrase to lookalike characters to see how a form or search box treats visually identical text.
Normalizing pasted usernames. Paste a name that contains Cyrillic or Greek lookalikes and get the plain Latin equivalent plus a list of the suspicious characters.
Spotting confusable text in a document. Run detection mode over a paragraph to find characters that are not the Latin letters they appear to be.

Before and after, using the standard map: Hello becomes Hеⅼⅼо (uppercase H stays Latin H because uppercase B–Z have no mapped lookalike; e maps to Cyrillic е U+0435, each l maps to ⅼ U+217C, and o maps to Cyrillic о U+043E), and normalizing Hеⅼⅼо returns Hello.

Tips for Best Results

Say the direction explicitly when your text already contains lookalike characters, since the agent cannot always tell intent from the text alone.
Keep the change report when you need to prove which characters were altered.
If a platform rejects the output, remember that homoglyph substitution changes code points, not the platform's rules.

FAQ

What input does it accept? Any plain text you paste or type. Very long text is handled in sections; you will be told if a section was cut.

Is the output exact? Yes for characters in the stated map. Unmapped characters are preserved exactly, and the report shows every substitution. The agent does not claim the result will pass any particular filter.

What makes this agent's output transparent? It lists its full map, states exactly which uppercase letters are unmapped, supports the reverse direction, and reports each change with code points. It does not produce font files, images or audio.