Skip to content

Text checks

AI Text Watermark Remover and Hidden Character Detector

Paste a passage and Slop or Not counts what came along with it: invisible Unicode, curly punctuation, and Cyrillic letters standing in Latin words. Every one is listed by Unicode name with its count, and a clean copy comes back in one click. The passage is processed in memory on Numen's server and discarded with the response, so the scan stores nothing. A hit will not tell you who wrote the text, because word processors leave the same marks. For a verdict on the writing itself, the AI text detector is a separate check, and it reads English only. It runs on Numen's private Mac server, away from outside AI companies.

AI Watermark Removers: What They Remove and What Survives

What it checks

See Every Hidden Character, Then Copy Clean Text

Invisible Unicode: zero-width spaces and joiners, word joiners, BOMs, and non-breaking spaces such as U+202F

Smart punctuation swaps: curly quotes, em dashes, ellipses, and fullwidth symbols

Lookalike letters from Cyrillic and Greek that pass for Latin characters

Good Uses

  • Checking a document for hidden characters before publishing or grading
  • Cleaning text pasted out of a chat window, PDF, or word processor
  • Verifying that copied text carries no invisible markers before sharing it

Limits to Know

  • Cleanup fixes mechanical fingerprints only; it does not alter statistical watermarks woven into word choices
  • Cleaned text is still the same writing; the AI text detector reads the words themselves
  • US/UK spelling conversion runs in the iPhone and Mac apps, not the online scan

How to check text for hidden characters

  1. 1

    Copy the text

    Copy the passage from Word, Google Docs, a PDF, or a chat window. Formatting does not matter; the scan reads characters, not styles.

  2. 2

    Paste it above

    Paste into the checker. The passage goes to Numen's server for the character scan, is processed in memory, and is discarded with the response.

  3. 3

    Scan and read the breakdown

    Press Find hidden characters. The breakdown counts every hidden character the scan found and lists each one by Unicode name with how many times it appears.

  4. 4

    Copy the cleaned text

    One click copies a version with the mechanical fingerprints removed. For a verdict on the writing itself, run the AI text check, a separate step that reads English only.

FAQ

Hidden Character Detector Questions

What are hidden characters in AI text?

Hidden characters are the ones you cannot see in rendered text: zero-width spaces and joiners, word joiners, byte-order marks, and unusual spaces such as the narrow no-break space at U+202F, which drew attention after readers spotted it in output from ChatGPT's o3 and o4-mini models. The scan also flags two visible-but-swapped families: smart punctuation such as curly quotes and em dashes, and Cyrillic or Greek letters shaped like Latin ones. None of them prove AI authorship on their own, because word processors, chat windows, and ordinary copy-paste produce them too.

How do I remove hidden characters from ChatGPT text in Microsoft Word?

Select the passage in Word, copy it, paste it into the checker above, then paste the cleaned version back over the original. Word will not show you what is there first. Its formatting-marks button reveals paragraph marks, tabs, and ordinary spaces, and stops well short of zero-width characters and lookalike letters. Find and Replace can pull out one code point at a time if you already know which to hunt, and the scan finds all of them in one pass. For a stack of papers, the iPhone and Mac apps run the same pass on each one offline, and on the Mac, Pro's slop cleanup command can script a whole folder.

How do I check a Google Docs document for hidden characters?

Select the text in Google Docs, copy it, and paste it into the checker above. Docs has no built-in view for zero-width characters or lookalike letters, and its own autocorrect adds curly quotes and em dashes as you type, so part of what the scan reports will be Docs' work rather than an AI tool's. Paste the cleaned text back to replace the original.

Can I check a PDF for hidden characters?

Yes, as long as the text is selectable. Copy the passage out of the PDF and paste it into the checker above; the scan reads characters, so page layout and fonts do not matter. A scanned PDF is really an image and carries no character data until you run OCR on it, and OCR output carries its own transcription errors rather than the original's hidden characters.

Does this remove AI watermarks?

It cleans mechanical fingerprints: invisible Unicode is removed, smart punctuation turns plain, and lookalike letters are swapped for their Latin twins. Non-breaking spaces are swapped for ordinary ones rather than deleted, so 100 % keeps its separation; check the spacing before you paste the clean copy over the original. It does not touch statistical watermarks, the kind Google's SynthID-Text weaves into the word choices themselves. Anthropic says future Claude models will carry a version of that same SynthID-Text mark. OpenAI's textGrain, announced on 5 October 2026, is a separate mark of the same type: it adds an invisible statistical signal to the model's word choices. A statistical mark survives a character swap, and reading one needs the provider's own detector. Cleaning also does not make AI writing read as human: the AI text detector reads the words, and it reads them the same way before and after.

What if no hidden characters are found?

That tells you the text carries no mechanical markers, and nothing more. Plenty of AI-generated writing has none, either because the tool never inserted any or because a trip through a plain-text editor stripped them, and plenty of human writing has some. For an answer about the writing itself, run the AI text check and read the verdict as supporting evidence rather than proof.

Does my text leave my browser?

Yes. The hidden-character scan and the cleaned output are produced in memory on Numen's server and discarded with the response, so the scan stores nothing. The AI text check is a separate step and travels to Numen's private Mac server, away from third-party AI providers. The iPhone and Mac apps run both checks entirely on-device.

Does removing hidden characters make AI text undetectable?

No. The AI text detector reads sentence structure and word choice, not invisible characters, so removing them does not change the evidence it reads. That model caught 99.5% of AI text at a 5% false-positive rate on RAID, a public benchmark we do not run, with 11 adversarial attacks in the test set. Treat any tool promising undetectable output after a Unicode strip as selling something it cannot deliver.

Can I clean text offline or from the command line?

Yes, on Apple devices. A stack of student submissions or client drafts never has to touch a network: the iPhone and Mac apps run the scan and the cleanup on-device, and they add spelling conversion between US and UK English that the online scan does not do. On the Mac, Pro also ships a slop cleanup command and an MCP server, so a whole folder can go through it from a script, or from Claude, Codex, or Cursor.

Why do AI tools leave hidden characters behind?

Mostly by accident. Chat interfaces render answers as rich text, so quotes curl, hyphens become em dashes, and typographic spaces such as U+202F get baked into whatever you copy out. Many documents also pick these characters up from a word processor or a website long before any AI touches them. A hit is the start of a question, not the answer to one.

What Unicode characters does the hidden character detector find?

The hidden character detector finds four kinds of character you cannot see: invisible characters such as zero-width spaces, joiners, and byte-order marks; directional (bidi) controls that change the reading order; Unicode tag characters, an invisible range that can carry hidden instructions into an AI prompt; and the two non-breaking spaces, U+00A0 and the narrow U+202F. It also flags two visible families: typographic punctuation such as curly quotes, em dashes, and ellipses, and Cyrillic or Greek letters shaped like Latin ones. Each one is listed by Unicode name with its count.

What does the AI hidden character remover change in my text?

The AI hidden character remover deletes invisible characters, turns non-breaking spaces into ordinary spaces, and replaces typographic punctuation with plain keyboard marks. Cyrillic and Greek lookalikes are swapped for the Latin letters they imitate. It never rewrites a word, so the cleaned copy reads the same as the original.

Is hidden text the same as hidden characters?

No. Hidden text, such as white-on-white or tiny type, or prompt-injection instructions tucked into a web page, is made of ordinary letters that styling hides. The scan reads characters, not styles, so it will not flag that text. Spotting styled hidden text takes a tool that reads the document's formatting.

Can a hidden character detector find an AI text watermark?

No. Statistical marks such as OpenAI's textGrain and Google's SynthID-Text live in the model's word choices, not in any one character, so a character scan cannot see them and cleanup leaves them in place. Reading one needs the provider's own detector; an AI watermark text detector that only checks characters will miss it every time. Our post on the ChatGPT watermark explains how textGrain works.

Run the Next Check on Your iPhone or Mac

The Slop or Not app checks images and text on-device, so nothing you check ever leaves your device.

Download on the App Store