Invisible Characters – Detect and Remove Hidden Unicode

What counts as invisible characters

The label covers a small family of Unicode code points that render to nothing in normal text. The detector recognizes the common offenders.

• Zero-width space – U+200B, the most-reported invisible marker.
• Zero-width non-joiner and joiner – U+200C and U+200D.
• Bidirectional controls – left-to-right and right-to-left marks like U+200E.
• Soft hyphen – U+00AD, a line-break hint that hides until needed.

Each of these invisible characters is valid Unicode, which is exactly why editors and forms rarely warn you when one slips in.

Why invisible characters cause problems

Because they have no visual footprint, invisible characters are debugged by frustration. The symptoms look impossible until you find the hidden code point behind them.

• A password "matches" by eye but is rejected on login.
• Code throws a syntax error on a line that looks perfect.
• A spreadsheet treats one column as two, or fails a lookup.
• A character count is higher than the text you can actually see.

Stripping the invisible characters resolves every one of these, because the parser and your eyes finally agree on what the text contains.

Invisible characters and AI watermarking

Several AI tools have experimented with embedding invisible characters into generated text as a covert signal. The technique works because, as detector vendor Originality.ai notes, the markers survive copy-paste and travel with the text even when it looks clean on screen.

The practice never became a standard, but enough early prototypes leaked that plagiarism and AI checkers now flag suspicious runs of invisible characters. Cleaning them before you publish removes that ambiguity, so your text is judged on its words rather than its hidden bytes.

How to remove invisible characters

The workflow is a single paste-and-scan loop. The detector does the locating; you decide whether to clean.

1. Paste the suspect text into the box on this page.
2. The tool scans and lists every hidden code point it finds.
3. Review the report to see where each one was located.
4. Remove them in one click and copy the cleaned text.

Run the check whenever you copy from a chat tool, a PDF, or a web page, since those are the most common sources of stray invisible characters.

When to keep invisible characters

Not every hidden code point is a mistake. Invisible characters do real work in the right context, so the detector reports before it removes.

• Right-to-left scripts rely on bidi marks for correct ordering.
• Soft hyphens give browsers good line-break points.
• Some emoji sequences use a joiner to combine glyphs.
• In plain English, almost any of these is safe to strip.

The rule of thumb: if you did not put the invisible characters there on purpose, removing them is the safe choice.