Invisible Character Checker
Find what you cannot see. Remove only what you choose.
This checks Unicode characters, not AI authorship or statistical AI watermarks. Hidden characters can be normal and meaningful in text.
No file selected
UTF-8 TXT or DOCX · up to 20 MB · 500,000 UTF-16 code units. Legacy .doc: save as .docx in Word first. DOCX imports the main document text only; headers, footers, comments, footnotes, images and formatting are not imported. Deleted tracked text is skipped. Results export as plain TXT, not Word.
Paste text or open a document to get started.
See hidden characters
| Unicode | Name | Count |
|---|
Character positions
Choose what to clean
All groups are kept by default. Removing joiners or variation selectors can break emoji or change spelling. We do not normalize accents, punctuation, tabs or ordinary line breaks.
All text and document processing happens locally. Only anonymous run IDs, durations and allowlisted error codes are sent for statistics. Text, documents and filenames are never uploaded.
How to use Invisible Character Checker
- Paste text or open a UTF-8 TXT or DOCX file.
- Scan the content and review the highlighted characters and their positions.
- Select only the character groups you want to remove, then copy or download the cleaned text.
What to know
Hidden characters do not prove that text was AI-generated. Some are necessary for languages or emoji. Convert legacy DOC files to DOCX first; exports are plain text.