Free AI content detector · no sign-up

Is your essayblog postpaperemailessay
written humanly?

Humanly scans your text sentence by sentence and shows you which parts read like AI, so you can see the why, not just a number.

0 / 2,000 words

[Confirm before launch] Your text isn’t saved after the check. Privacy Policy

More than a score.
A reason.

Built for students, teachers, writers and editors who want to understand a result, not just accept it.

Sentence level

See exactly which sentences look machine-made

Every sentence gets its own probability, so you know what to revise.

Private by default

[Confirm your policy] Your text isn’t saved after the check and is never used for training.

Reports you can share

Coming soon: download a PDF or send a link to a teacher, editor or client.

Nothing to install

Works in your browser on desktop, tablet and phone.

Three steps.
About ten seconds.

01

Paste your text

Up to 2,000 words per free check, no sign-up needed.

02

We scan it

Our model reads the whole text and every sentence, on its own and in context.

03

Revise with clarity

See the flagged sentences and rewrite them in your own voice.

Pick your plan

FREE

£0
Free forever, no card needed
10,000 words / month
  • Overall AI probability score
  • Sentence-level highlights
  • Up to 2,000 words per check
  • No sign-up required
Start for free

PREMIUM

MOST POPULAR
£12.99/ month£23.99
Billed £155.88 per year
300,000 words / month
  • Everything in Free
  • Up to 5,000 words per check
  • Coming soon: file uploads, PDF reports, Plagiarism Check
Choose Premium

PROFESSIONAL

£24.99/ month£45.99
Billed £299.88 per year
500,000 words / month
  • Everything in Premium
  • Up to 5,000 words per check
  • Priority email support
  • Coming soon: batch scanning
Choose Professional

Prices include VAT [confirm]. Cancel anytime.

Heard from humans

Write a review
[Star rating]

[Paste a real review here, with the reviewer’s permission.]

[Reviewer name] · [Source]
[Star rating]

[Paste a real review here, with the reviewer’s permission.]

[Reviewer name] · [Source]
[Star rating]

[Paste a real review here, with the reviewer’s permission.]

[Reviewer name] · [Source]
[Star rating]

[Paste a real review here, with the reviewer’s permission.]

[Reviewer name] · [Source]

Questions,
answered.

Something else? Email us.

What is an AI detector?

An AI detector estimates how likely it is that a piece of text was written by an AI model such as ChatGPT. It looks for statistical patterns that AI writing tends to have, like very predictable word choices and uniform sentence rhythm. Humanly also highlights the sentences behind the score, so you can see why a text was flagged instead of getting a single number.

Is Humanly really free? Do I need an account?

Yes. The Free plan gives you 10,000 words a month with no sign-up and no card needed. Premium (300,000 words a month) and Professional (500,000 words a month) give you more words and longer checks; sign in with Google to subscribe. Paying yearly saves 45%. You can cancel any time from your account.

How accurate is Humanly’s AI detection?

No AI detector is 100% accurate, and anyone who claims otherwise is overselling. Humanly gives a probability, not a verdict. Accuracy is highest on longer texts (150+ words) written in English. [Publish your measured accuracy and false-positive rate here, with a link to how you tested it.]

Can my own writing be flagged as AI?

Yes, it can happen. Very formal or formulaic writing, text written by non-native English speakers, and short passages are more likely to be misread as AI. If you wrote it yourself and it’s flagged, keep your drafts and notes as evidence, and treat the score as one signal among many.

How does Humanly tell human and AI writing apart?

Humanly uses a machine-learning model trained on large collections of human-written and AI-generated text. It scores the whole document and each sentence, reading every sentence both on its own and alongside its neighbours, and highlights the parts that look most machine-written.

Which AI models can it detect?

[List the models you have tested against, e.g. ChatGPT, Claude, Gemini, Llama, DeepSeek.] New models are released often, so detection of the very latest ones may lag behind until the detector is updated.

Can it detect AI text that has been paraphrased or edited?

Partly. Light edits usually still show AI patterns, but heavily rewritten or mixed human-and-AI text is harder to detect and more likely to get a medium score. The sentence highlights help you see which parts still read as machine-written.

Which languages can I check?

English only for now. Results for other languages are not reliable, so please don’t rely on them.

How long does a check take?

Most checks finish in a few seconds. Very long documents can take a little longer.

Is my text safe? Do you store it?

[Confirm your real policy before launch, e.g.: Your text is sent over an encrypted connection, is not saved after the check, and is never used to train our models.]

How do I get the most accurate result?

Check at least 150 words of continuous English text. Remove headings, reference lists, quotes and code, which can skew the score. For long papers, checking sections one at a time often shows more clearly where the AI-like passages are.

My essay got a high AI score. What should I do?

Look at the highlighted sentences first. Rewrite them in your own voice, with specific examples, your own opinions and details only you would know.

Should teachers use Humanly as proof of cheating?

No. An AI score is not proof that a student used AI. Use it to start a conversation, and look at other evidence such as drafts, version history and the student’s usual writing style before drawing any conclusion.

Do I need to install anything?

No. Humanly runs in your browser on desktop, tablet and phone.

How can I get help if something goes wrong?

Email us at [SUPPORT EMAIL]. We usually reply within [RESPONSE TIME].

Find out in seconds.

10,000 free words every month. No sign-up.

Scan my text