Text checks
ChatGPT Watermark Remover
Paste a ChatGPT answer and get it back clean. Slop or Not cleans up what rides along when you copy from a chat window: invisible Unicode goes, curly quotes and em dashes turn plain, and the narrow no-break space that looks like an ordinary one becomes an ordinary one. Every word stays where it was. The same check tells you how likely it is that AI wrote the text. One thing no remover can strip is textGrain, the watermark OpenAI announced on 5 October 2026. It lives in which words the model picked, not in the characters. Text is checked on Numen's own servers, away from outside AI companies, and deleted after processing.
Catches 99.5% of AI text at a 5% false-positive rate, scored by RAID, the public benchmark, not by us.
Paste a paragraph or a whole thesis. The iPhone and Mac apps have no word cap and no monthly word budget.
- Paste English textSchool essays, product reviews, captions, articles, and social posts up to 20,000 characters.
- ResultA probability verdict, AI score, and readability analysis.
- PrivacyDeleted the moment the check finishes. We never keep your text.
What it checks
Hidden Characters Come Off Here. The Watermark in the Words Does Not.
Invisible Unicode such as zero-width spaces, joiners, and byte-order marks, taken out of the clean copy, and lookalike Cyrillic and Greek letters swapped for Latin ones
Curly quotes, em dashes, and ellipses turned into plain punctuation, and the U+202F narrow no-break space turned into an ordinary space
English writing tells highlighted in the passage, next to the verdict from the RAID-scored AI text detection model
Good Uses
- Pasting a ChatGPT draft into a newsletter, a help-center article, or a slide deck without the curly quotes and invisible characters it carried out of the chat
- Removing zero-width characters from ChatGPT-written text before it goes into code, a spreadsheet, or a form field that trips over them
- Seeing what a ChatGPT-assisted draft carries before you publish it: hidden characters, writing tells, and the AI verdict side by side
Limits to Know
- Cleanup cannot read or remove textGrain, which OpenAI built into the word choices: detecting it takes OpenAI's secret key, and the AI verdict cannot tell whether it is there
- Writing tells and the AI verdict read English only; cleanup works on text in any language
- Non-breaking spaces, U+202F included, become ordinary spaces rather than being deleted, so check the spacing in numbers and French punctuation before you paste the clean copy back
How to remove ChatGPT watermark characters from text
1
Paste the text and run the check
Copy the answer from the ChatGPT window, paste it into the checker above, and run the check. The AI text detection model reads the writing, and the character scan is processed in memory on Numen's server and discarded with the response.
2
Copy the cleaned text
The clean copy comes at the end of the result. One click puts it on your clipboard with the invisible characters gone, lookalike letters swapped for Latin ones, and the punctuation plain. If the scan finds nothing, it says so, and the text you pasted is already the clean copy.
3
Read the writing tells
If the text is English, the tells panel marks the clichรฉs and sentence patterns that turn up again and again in AI writing. They are worth a look, but read them as evidence about the writing, not as a score to chase.
FAQ
ChatGPT Watermark Remover Questions
What does the ChatGPT watermark remover take out?
It takes out invisible Unicode such as zero-width spaces and joiners, word joiners, and byte-order marks, and it swaps Cyrillic or Greek letters posing as Latin ones for the Latin originals. Curly quotes, em dashes, and ellipses turn into plain ASCII, and non-breaking spaces, including the narrow one at U+202F, turn into ordinary spaces. No word is rewritten, so what you paste back says exactly what it said before.
Does it remove the textGrain watermark?
No. OpenAI says textGrain adds an invisible statistical signal to the model's word choices, so the mark lives in which words were picked, not in any character on the page, and no character cleaner, ours included, touches it. OpenAI said textGrain would reach eligible ChatGPT and Codex text in the EU over the coming weeks, and that API customers around the world could opt in on select models from 5 October 2026. Reading it takes OpenAI's secret key, and OpenAI said access to its detector would initially be limited to approved researchers and expert organizations. Slop or Not's verdict reads the writing, not the watermark, so it cannot tell you whether a passage still carries textGrain.
Does editing ChatGPT text remove the watermark?
According to OpenAI, changing the words weakens detection, and OpenAI also says a missing watermark does not prove a human wrote the text. Cleanup changes no words, so it does nothing to the watermark. The AI text detection model is a separate check, scored on the RAID test set, whose attacks include paraphrasing.
Can I clean text without running an AI check?
Yes. If a clean copy is all you want, the hidden character detector is the better tool. It runs the same cleanup without an AI verdict, so it does not spend one of your free daily checks or wait in the detection queue, and it reads text in any language.
What happens to the text I paste?
The character scan and the cleaned copy are produced in memory on Numen's server and discarded with the response, so the passage is never stored. The AI text check is a separate step and travels to Numen's private Mac server, away from third-party AI providers, and the text is deleted after processing. The iPhone and Mac apps run both checks on-device.
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.
More from Numen
Two more apps that run on your device. Neither needs an account.
Private LLM
A Local AI chatbot for uncensored open-source models that runs offline on iPhone, iPad, and Mac. Private LLM uses OmniQuant and GPTQ quantization tuned per model, so its 3-bit models match the 4-bit models other apps ship.
Clean Links
Most tracking now hides behind link redirects. Clean Links unwraps short links and strips every tracking parameter from a URL before you share it, on iPhone, iPad, Mac, and the web.
Related checks
Compare Related Detector Workflows
ChatGPT Watermark Detector
Check text for the hidden characters ChatGPT leaves behind, like the U+202F narrow no-break space, and read the writing itself with a RAID-scored detector.
Hidden Character Detector
Free hidden character detector: scan text for invisible Unicode, smart punctuation, and lookalikes, then copy clean text; Numen never stores the passage.
AI Text Detector
Check whether writing looks AI-generated with Slop or Not. Paste text online for a quick check, or use the iPhone and Mac apps on-device.