Supported hidden markers
None foundPaste text to scan it locally.
Free browser-local tool
Remove zero-width and invisible characters
Invisible Unicode characters can enter text through copying, formatting, tracking, bidirectional layout, AI chat output, or deliberate obfuscation. This free local tool makes supported zero-width characters, Unicode tag characters, bidi controls, byte-order marks, word joiners, and unusual space characters visible in a category report before cleaning. You can copy the normalized result and see the count removed from each category. Ordinary letters, punctuation, line breaks, and emoji are preserved; zero-width joiner and non-joiner characters are kept by default because they can be required for emoji sequences and correctly written languages. The free cleaner does not judge whether text was written by a person or AI, bypass a detector, upload the text, or rewrite its meaning. The separate paid AI Rewrite step does rephrase your text on a signed-in server task. Neither step claims to detect or remove a statistical token-distribution watermark such as the one described for Claude.
Paste text to inspect hidden characters
Find supported zero-width characters, Unicode tags, abnormal spaces, bidirectional controls and invisible text markers.
AI rewrite · 1 credit
Server task · credits requiredRewrite with AI
Uses a paid provider to rephrase copied or AI text with clearly different wording. Sign in to use credits; the free cleaner above needs no account.
Add text in the editor above before rewriting. Check “I wrote this text or have permission to rewrite it” to continue.
AI rewrite results and billing receipts are saved to your account so interrupted requests can be recovered. The free cleaner above remains browser-local.
Rewriting changes your wording and is best-effort: it is not a detector, does not guarantee removal of any token-level watermark, and must not be used to pass authorship, fraud, or platform checks.
Supported removal types
Supported invisible text categories
The cleaner recognizes a deliberately narrow set of code points so its report is explainable. These include zero-width space, word joiner, embedded byte-order marks, Unicode tag characters, supported bidirectional embedding or isolate controls, and non-standard spacing characters that can be normalized to an ordinary space. Counts are shown before any change is applied.
Zero-width joiner and zero-width non-joiner are different: they can shape words in some writing systems and combine emoji into a single visible glyph. The safe preset therefore keeps them. The interface should not call all invisible code points malicious, because direction controls, variation selectors, combining marks, and joiners can have legitimate linguistic or typographic roles.
- U+200B zero-width space, U+2060 word joiner, and embedded U+FEFF
- Unicode tag characters in the supported U+E0000–U+E007F range
- Supported bidi embedding, override, and isolate controls included by Text Clean
- No-break, thin, narrow, ideographic, and other supported unusual spaces normalized by the preset
How it works
Inspect code points before copying cleaned text
The text remains in browser memory; the tool needs no file upload, account, or network processing.
- 01
Paste or type text into the input. Nothing is removed while you are still reviewing it.
- 02
Inspect the category counts for the supported hidden characters found in the input.
- 03
Review the fixed Text Clean preset, including its supported bidi and spacing rules, and keep the original when those code points may be meaningful.
- 04
Run the clean and compare the category totals, line breaks, and visible text.
- 05
Copy the result. Keep the original when working with legal, multilingual, signed, hashed, or byte-sensitive text so you can verify intended meaning and direction.
Metadata vs pixels
Invisible text is not image metadata or a pixel watermark
A hidden Unicode character is part of a text string. It can affect searching, comparison, cursor movement, line wrapping, direction, or rendering even though it has little or no visible width. Cleaning it changes the sequence of code points, not metadata attached to a separate file. AI chat output and copied text can carry these markers alongside the visible answer.
Image metadata lives in EXIF, XMP, IPTC, C2PA, or format-specific chunks, while an invisible image watermark is encoded into pixels. This text page handles neither. The watermark Anthropic has described for Claude is a statistical pattern across generated tokens, not a Unicode character inserted into the output; a character cleaner cannot detect or prove the removal of that fingerprint.
What is preserved
What the safe text preset keeps
Visible letters, digits, punctuation, symbols, tabs, and line breaks remain in order. The cleaner keeps zero-width joiner, zero-width non-joiner, variation selectors, combining marks, and ordinary directional characters by default because automatic deletion could break scripts, emoji, names, or technical content.
Cleaning does not paraphrase, spell-check, translate, summarize, or alter letter case. Supported unusual spaces are normalized to ordinary spaces and reported separately. The output report shows category and removal totals so a user can notice an unexpected difference before copying.
Limitations
Language, direction, and security limits
Removing bidi controls can change the displayed order of mixed right-to-left and left-to-right text. Normalizing spaces can alter line wrapping or byte-for-byte signatures. Joiners and selectors may be invisible but semantically important. Review multilingual, source-code, domain-name, legal, and cryptographically signed material with the appropriate specialist tools before replacing the original.
This cleaner can expose and remove supported characters, but it is not a malware scanner, homoglyph detector, plagiarism checker, AI detector, or authorship verifier. It does not promise that cleaned text is safe, original, human-written, or accepted by a platform.
Reproducible example
Test-fixture example: category counts and a stable visible result
A literal Unicode fixture makes every expected change reproducible without using private user text.
Input
- The sequence “alpha<U+200B> beta<U+2060> gamma” with escaped markers replaced by the actual code points
- One Unicode tag sequence, one narrow no-break space, one bidi isolate pair, and the emoji family sequence joined with U+200D
- Expected counts recorded for zero-width, tag, bidi, unusual-space, and protected-joiner categories
Change
- The Text Clean preset removes U+200B, U+2060, and the tag sequence and normalizes the supported unusual space.
- Bidi controls are removed only in the explicit bidi test; U+200D joiners remain in every default-preset run.
Verification
- The output equals the fixture's expected code-point sequence, not merely the same-looking rendered text.
- Category totals match the number of removed, normalized, and preserved characters.
- The joined emoji remains one intended sequence and ordinary words and line breaks remain unchanged.
FAQ
Questions about this tool
Why does the tool keep some invisible characters?
Joiners, variation selectors, combining marks, and direction controls can be meaningful. Text Clean removes only its documented supported set and reports the categories before you replace the original.
Can this clean text copied from a PDF or website?
Yes, if you paste the extracted text. It cleans the string in the input box; it does not edit the original PDF, webpage, document, or image.
Does cleaning text from Claude or another chatbot remove its watermark?
Text Clean removes the supported hidden characters that can appear in copied output. Anthropic's described Claude watermark is a statistical pattern across the token distribution rather than an inserted character, so this tool cannot detect or prove its removal. The paid AI Rewrite changes the wording itself as the only known practical intervention; see the Claude text watermark guide for the full distinction.
How does the paid AI Rewrite work here?
Text Clean includes a signed-in AI Rewrite step that uses 1 credit to rephrase your text through a separate provider with different wording and structure. New accounts receive 3 welcome credits; a confirmed provider failure returns the credit. Rewriting is best-effort and is not a detector or a guarantee.
Does cleaning make text pass an AI detector?
No. The tool does not evaluate authorship or optimize for detectors. It only reports and changes the Unicode code points and spacing covered by its fixed preset.
Is the text sent to a server?
The free cleaner runs in the current browser and does not send the entered text to perform the operation. Clear the input when you finish on a shared device.
Related tools and guides
Continue with a verified path
Ownership notice
Use the text cleaner only on material you are permitted to edit. Do not use hidden-character removal to alter evidence, signatures, official records, another person's statement, or content for fraud or impersonation.