Writing tells

AI Words Checker for Clichés and Writing Tells

Paste English text and run the check. The AI words checker marks 64 familiar moves, from stacked questions to "it is important to note," exactly where they appear. Open the match list to jump back to any sentence. Nothing it finds changes the AI verdict. The phrase scan, hidden-character cleanup, and the AI check are separate tools, and each one sends your text to a Numen server.

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.
PrivacyChecked on Numen's private Mac server and never passed to an outside AI company. For text that stays on your device, use the app.

What it checks

What the AI Words Checker Marks in Your Text

The exact stock phrase, with its surrounding sentence

Runs of questions or repeated sentence openings

Leftovers such as "as of my last update"

Good Uses

Spotting habits in a draft before publishing
Showing an editor why a sentence was flagged
Reading where the stock phrasing sits alongside the AI verdict

Limits to Know

A person can write any of these phrases
The matcher works in English and matches text, not meaning
A fixed list can miss AI text and catch ordinary prose

FAQ

AI Words Checker Questions

What does the AI words checker flag?

It looks for a fixed list of 64 phrases and sentence shapes. They include stacked questions, repeated openings, stock transitions, promotional language, and copied chatbot leftovers. The highlight stays on the words that triggered it, so you can judge the sentence yourself.

Does a writing tell prove that text is AI-generated?

No. People use these phrases too. A match points to a line worth reading again. It never changes Slop or Not's AI verdict.

Does my text leave the browser?

Yes. The phrase scan runs in memory on Numen's server and is discarded with the response, so the scan stores nothing. Hidden-character cleanup works the same way. The AI check is a separate step and travels to Numen's private Mac server.

How is this different from the AI text detector?

This checker points to words from a fixed list. The AI text detector reads the passage as a whole and returns a probability. One can find nothing while the other scores high, or the reverse. Neither proves who wrote it.

How long does my text need to be for the density and baseline figures?

The density figure needs 80 words. Below that, one match alone reads as over one per 100 words, so the figure would report the length of the passage instead of the writing in it. The reference comparison needs 968 words. That reference is a fixed sample of pre-2023 books, news and documentation, and it scores 0.1 matches per 100 words. Under 968 words a single match clears it on arithmetic alone, which tells you nothing. Short passages still get the highlights and the list of matches.

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

Three more apps that run on your device. None 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.

Private Diffusion

Run the latest frontier image models on your iPhone, iPad, and Mac. Private Diffusion trains each one with quantization aware distillation (QAD), so the small model keeps the quality of the full size one.

Related checks