What is the difference between invisible characters and zero-width characters?

Zero-width characters (U+200B, U+200C, U+200D, etc.) are a specific subset of invisible characters that have an explicit width of zero. Invisible characters is a broader term that also covers directional marks, non-breaking spaces, control characters, and irregular whitespace — all of which render without a visible glyph in standard text flow.

How can invisible characters end up in my text?

The most common sources are: copying text from web pages (which embed directional and formatting marks), AI tool exports (which add watermarking characters), word processor conversions (which introduce non-breaking spaces and soft hyphens), and PDF text extraction (which adds spurious spaces and control characters).

Can invisible characters cause real problems?

Yes. Non-breaking spaces cause word-wrap failures in email clients and CMSs. Zero-width characters fragment tokens in search indexes and NLP models. Directional marks can reverse text rendering direction in bidi-naïve applications. Control characters can corrupt database fields or break JSON parsing.

Will this tool accidentally remove legitimate invisible characters?

The tool removes characters that have no semantic role in standard editorial text. If you are working with scripts like Arabic, Hebrew, or Khmer where U+200C and U+200D control ligature rendering, use the diff view to verify the output carefully before exporting.

Is there a copy-paste trick to generate invisible characters?

Yes — the internet is full of 'invisible text copy' pages that generate U+200B or U+3164 (Hangul Filler) for use in usernames and messages. This tool detects and removes all such characters regardless of their source.

Free AI Text Cleaner

Invisible Character Remover

Detect and remove all invisible characters from any text — zero-width spaces, directional marks, non-breaking spaces, control characters, and every hidden Unicode artefact. Free, private, browser-based.

Paste Your Text

Characters 0/10,000

Supports text from ChatGPT, Claude, and other AI assistants

Your text is processed locally. We do not store your data.

Upload Files

Drag & drop files or

Browse Files

Supported: .txt, .docx, .pdf (max 10MB, up to 10 files)

Cleaning Settings

Original Text

0 characters

Paste or type text to analyze...

Cleaned Text

0 characters

Cleaned text will appear here...

Your text is processed locally. We do not store your data.

Invisible characters are a broad category of Unicode code points that produce no visible glyph in standard text rendering. They enter text through copy-paste from web pages, AI tool exports, word processor formatting, and deliberate watermarking. This tool performs a comprehensive scan across all Unicode categories of invisible characters and removes every one, leaving clean, portable, standard text.

All invisible character types this tool removes

Invisible characters fall into several Unicode categories. This tool targets all of them in a single pass.

Zero-Width Characters

U+200B–U+200D, U+FEFF

Zero-Width Space, Non-Joiner, Joiner, and the Byte Order Mark. The most common AI watermarking characters. Invisible in all editors and browsers but detectable by parsers.

Directional Marks

U+200E / U+200F

Left-to-right and right-to-left marks used for bidi text control. Invisible in monolingual text but present in AI exports of mixed-language content and some copy-paste from web pages.

Non-Breaking Space

U+00A0

Looks identical to a regular space but breaks word-wrap differently and is treated as a distinct token by NLP models and databases — a reliable AI-text indicator when overused.

Soft Hyphen

U+00AD

An optional line-break hint embedded inside words by some AI tools (notably Gemini). Invisible in reading mode but fragments words in search indexes and NLP tokenisers.

Format & Control Characters

U+2060–U+2064, U+FFF9–U+FFFB

Word Joiner, Invisible Separator, Interlinear Annotation characters. Rare in legitimate text; their presence almost always indicates a programmatically generated source.

Irregular Whitespace

U+2000–U+200A, U+3000

En space, em space, thin space, ideographic space, and other typographic spaces. Visually similar to a regular space but distinct at the Unicode level, causing issues in databases and search.

How to remove invisible characters in 3 steps

1
Paste or upload
Paste text directly or upload a .txt, .docx, or .pdf file. Up to 10 files can be queued for batch cleaning.
2
Full-spectrum invisible-character scan
The engine checks every character against a comprehensive Unicode blocklist covering all invisible, zero-width, directional, control, and irregular-whitespace code points.
3
Review & export
The diff view shows every removed character with its Unicode code point. Download the clean file or copy the text. The stats panel confirms exactly what was removed.

Before & after invisible character removal

Before — contains hidden markers

Copy-paste from‎ web‏ pages⁠ often brings hidden invisible characters.

After — clean, standard Unicode

Copy-paste from web pages often brings hidden invisible characters.

Frequently asked questions about Invisible Characters watermarks

What is the difference between invisible characters and zero-width characters?: Zero-width characters (U+200B, U+200C, U+200D, etc.) are a specific subset of invisible characters that have an explicit width of zero. Invisible characters is a broader term that also covers directional marks, non-breaking spaces, control characters, and irregular whitespace — all of which render without a visible glyph in standard text flow.
How can invisible characters end up in my text?: The most common sources are: copying text from web pages (which embed directional and formatting marks), AI tool exports (which add watermarking characters), word processor conversions (which introduce non-breaking spaces and soft hyphens), and PDF text extraction (which adds spurious spaces and control characters).
Can invisible characters cause real problems?: Yes. Non-breaking spaces cause word-wrap failures in email clients and CMSs. Zero-width characters fragment tokens in search indexes and NLP models. Directional marks can reverse text rendering direction in bidi-naïve applications. Control characters can corrupt database fields or break JSON parsing.
Will this tool accidentally remove legitimate invisible characters?: The tool removes characters that have no semantic role in standard editorial text. If you are working with scripts like Arabic, Hebrew, or Khmer where U+200C and U+200D control ligature rendering, use the diff view to verify the output carefully before exporting.
Is there a copy-paste trick to generate invisible characters?: Yes — the internet is full of 'invisible text copy' pages that generate U+200B or U+3164 (Hangul Filler) for use in usernames and messages. This tool detects and removes all such characters regardless of their source.

Clean text from other AI tools

Zero-Width Remover

Focused removal of U+200B, U+200C, U+200D, U+FEFF, U+00AD and all other zero-width Unicode characters.

ChatGPT Remover

Removes the specific invisible character patterns and HTML data attributes embedded by ChatGPT.

Or return to the main watermark eraser to clean text from any AI tool.