how it works
How Neuroslop tells living text from machine text
In short
Neuroslop is a detector of careless machine text and a humanizer in one. It scores how alive a text sounds from 0 to 100 and shows every suspicious spot: not a guessed “AI probability”, but concrete findings with an exact address. Checking is free and unlimited, and we never store your text. And we're no “AI police”: well-written text passes even if a model typed it. What gets caught is sloppy work.
Who's behind it
Neuroslop is an independent project by a small team obsessed with one thing: telling living writing from machine filler. We build the detector, the humanizer, the bot and the API ourselves and improve them daily. No venture pressure, no data resale — just a tool we want to use ourselves. Questions, ideas or partnership? The contacts page lists every way to reach us.
What the detector looks for
- AI clichés. Hundreds of signature phrases models repeat from text to text — from “in today's world” to empty intensifiers and crutch connectors. The dictionaries grow daily.
- Obsessive repetition. A word that follows the reader around. We tell lazy repetition from terms and proper names — an article about one company gets no false penalties.
- Dead rhythm. Sentences of identical length marching in a row — the signature gait of generation. Living text breathes: short, then long.
- Structural signatures. Stamped antitheses, lists of exactly three items, machine punctuation. Harmless one by one — together they give the author away completely.
Why we don't accuse the innocent
- Genre first, verdicts second. The detector understands what it's reading — news, a contract, an academic paper or a personal post — and judges each by its own standards. Dry business prose is not slop.
- One marker is not a verdict. Only the weight of evidence decides. A single unlucky word will never sink a text.
- Checked against the classics. The dictionaries are verified on a corpus of real literature, so human turns of phrase never become false markers.
- Formal style is not a crime. Ordinary detectors are notorious for falsely accusing students and lawyers of being robots. We built our method so a careful formal writer always passes.
A detector that learns every day
User reports, fresh AI writing habits, our own research — all of it feeds the dictionaries in all 20 languages, daily. How exactly this kitchen works is our trade secret: we keep the recipe unpublished precisely so it can be neither copied nor gamed.
Humanization
Our proprietary humanization algorithm rewrites the text: it removes the flagged clichés, breaks the monotone rhythm, and keeps the meaning, genre and the author's voice. Humanization is not neutralization: emotional text stays emotional. The before/after score shows instantly. Checking is free; humanization gives 3 free rewrites a day, then credit packs — Standard and Premium.
Languages
The interface and checking work in 20 languages. Russian and English get the deepest analysis, and every language has its own cliché dictionaries — not machine-translated English rules.
Where it works
- The website: free checking and a workspace with humanization.
- The Telegram bot @AITextChecker_bot: the same core and plans, paid with Telegram Stars.
- A public HTTP API for developers: check and humanize by key.