is it slop?

Is it slop? An AI text detector for checking whether ChatGPT or another model wrote a passage.

0 words

Preview, unlimited checks

Is it slop?

Paste it, and I'll mark it up.

  • A blunt call, or an honest "too close to say"
  • How sure I actually am, a number rather than a vibe
  • The signals behind it, so you can weigh them yourself.

Don't fail a student or pile on a stranger over a score. Detectors can get it wrong, so read the signals and use your own judgement.

How this works

  1. You paste a passage

    A paragraph or more, up to 1,000 words. The more words I have, the more confident I can be.

  2. I run the checks

    I look for what ChatGPT and other AI models leave behind in the text. The telltale smells.

  3. You get the call and the working

    A blunt call, how sure I am, and what the check actually measured.

Know more about slop and detection

What does "slop" mean here?

Slop is content that is made with AI with little or no human effort or care. Usually AI slop isn't worth your time and might contain hallucinations or flat-out lies. This AI slop detector helps you find out where the text comes from, not if it's good or not. Can humans write slop as well? Yes, but we only focus on finding the slop created with AI.

Are AI detectors accurate?

AI detectors are not 100% accurate. Yes, even this one. AI detectors try to find patterns, styles and predictability in generated text. Content that is well structured or edited by professional writers or editors can sometimes trip them. Treat a high score as a signal to be suspicious, or to see if it's worth your time and attention.

Can a teacher tell if I used ChatGPT?

Yes, if you just copy/paste the AI text without making it your own. A detector can say a passage looks AI-written, but not who actually wrote it. People use AI to improve their writing, and this tool will help you identify if your text is more human-like or machine-like.

What does the score actually mean?

It is a likelihood. Eighty per cent means the text contains a lot of things that are present in AI output. The highlights show which parts we think are generated by AI when the text has mixed authorship. When the number of words in a text is low, we say the confidence is low as well.

Which AI models can it spot?

It looks for patterns that any AI model leaves behind, not one in particular. All models are predictive text generators underneath, and ChatGPT, Claude, Gemini, Grok and others all leave telltale signs. AI text a person has rewritten by hand is the hardest to detect. If they did, we wouldn't call it slop anyway.

Can it check images, files or a URL?

Not yet. We are building these features and they will be available soon.

Is it free?

For now, yes: it is in preview and checks are unlimited. A paid tier with a free allowance is planned.