Skip to content
Free · Unlimited checks

An AI detector that shows its work.

Find out whether your text reads as AI-generated, which sentences raised the score, and why — in seconds.

Unlimited on every plan, including Free.
PCOS Literature Review18% AI

Polycystic ovary syndrome affects a substantial share of women of reproductive age, yet the criteria used to diagnose it remain contested. It is important to note that the heterogeneity of PCOS presents significant challenges for researchers and clinicians alike. Most studies relied on the Rotterdam criteria. This review synthesizes current evidence on diagnostic approaches.

Why it was flaggedFormulaic opener“It is important to note that” is a stock phrase common in AI-generated prose.
What you get

More than a percentage.

Overall AI score

One clear probability with a human / AI split, so you know where the draft stands.

Sentence highlights

Passages that raised the score are highlighted, with weak signals marked separately.

Why it was flagged

Click any highlight for the reason and a concrete suggestion for what to change.

Stock-phrase checks

Formulaic openers, stacked transitions and overused words are pointed out by name.

History

Recent checks are kept in your browser so you can come back and compare drafts.

One click to rewrite

Send flagged text straight to the Humanizer when you want to revise it.

How it works

How our AI detector decides.

Every result combines a document score, per-sentence scores and pattern checks. You see all of them, not just the total.

01

A trained classifier

A transformer model fine-tuned on large sets of human and AI-written text. Not a chatbot asked to guess.

02

Sentence by sentence

Each sentence is scored with its neighbours for context, so you see where the signal comes from.

03

The whole document

Long texts are scored in overlapping sections and combined into one overall probability.

04

Writing patterns

Stock phrases, filler transitions and overused words are flagged by name, with what to write instead.

Know the limits

When to be careful with a score.

No detector is perfect, ours included. Results are less reliable in some situations, and a score should never be used on its own to accuse anyone. Read more in how AI detectors work.

Short texts. A few sentences rarely carry enough signal to judge.
Formulaic genres. Abstracts, legal text and templates are naturally predictable.
Non-native writers. Simpler, more regular phrasing can raise a score unfairly.
Heavily edited drafts. Mixed human and AI writing can land anywhere on the scale.
Use cases

Check before you publish, submit or send.

Bloggers & content writers

Edit AI-assisted drafts so articles read like you wrote them, before they go live.

Academics & researchers

Check papers and reports, and keep technical terms, citations and figures untouched.

Professionals

Polish emails, proposals and reports into clear, confident language in the tone you need.

Personal use

Cover letters, applications, personal statements: make everyday writing sound like you.

Want to fix what was flagged? The AI Humanizer rewrites it into natural language.

FAQ

AI Detector questions.

Can’t find what you need? Visit the Help center or contact us.

Is the AI Detector really free?

Yes. AI detection is unlimited on every plan, including Free. You only need a free account.

How accurate is it?

Detection is probabilistic and no tool is always right. That’s why we show the flagged sentences and the reasons behind every score. See our Disclaimer.

Which languages are supported?

English. Results on other languages are not reliable yet.

How much text can I check at once?

Up to 10,000 characters per check on Free, and 50,000 characters on paid plans. Longer documents are scored in overlapping sections.

Can I use a score to prove someone used AI?

No. A score is a signal to look closer, not evidence of authorship.

Paste your draft. See what reads as AI.

Free to start, right in your browser.

Check text free