Can it detect ChatGPT, Claude and Gemini?

Yes — the detector is model-agnostic. It scores statistical properties of the text rather than looking for a specific vendor's fingerprint, so it flags output from ChatGPT, Claude, Gemini, Copilot and open-weight models alike.

Does it work on documents that were AI-assisted rather than AI-written?

Yes. Sentence-level scoring shows mixed documents as a patchwork — high-confidence AI passages sitting alongside clearly human ones — which is usually the signal you actually need.

Is the detector free?

Your first check is free with no signup. Heavier use is covered by the £9.99 24-hour Day Pass or the £39.99/month Professional plan.

What happens to the documents I upload?

Every uploaded file is permanently deleted from DocumePro servers within 60 minutes under our Non-Persistence Guarantee. Nothing is retained for analytics, shared with third parties, or used to train AI models.

AI content detector — spot ChatGPT and Gemini-written text

Paste or upload any CV, cover letter, witness statement or report and get a confidence score for AI authorship, sentence-level highlighting, and a plain-English explanation you can share with a hiring manager, tribunal or auditor.

How the detector works

The detector scores the text on perplexity and burstiness — how predictable each token is, and how much sentence length and structure vary. Human writing is uneven; large language models produce unusually smooth, evenly-weighted prose. The model is tuned on UK English so British spelling and formal legal register do not trigger false positives.

Each sentence is scored individually, so you can see whether a whole document was generated or whether a human draft was polished by AI — the far more common pattern in job applications and consultancy reports.

Using it fairly

AI detection is probabilistic. Treat the score as a triage signal that tells you which documents deserve a closer human read, an interview question, or a request for the working draft. It should never be the sole basis for rejecting a candidate or disciplining an employee, and UK employers relying on it should record a human decision step.

For recruitment workflows, pair the detector with CV Intel so a flagged CV is still scored on substance against the job spec rather than being discarded automatically.

Typical uses

Screening CVs and cover letters for wholly AI-generated applications
Checking whether a submitted witness statement was drafted by a model
Verifying that outsourced content or consultancy reports were genuinely written
Auditing student or trainee submissions before a competency sign-off

Frequently asked questions

Can it detect ChatGPT, Claude and Gemini?: Yes — the detector is model-agnostic. It scores statistical properties of the text rather than looking for a specific vendor's fingerprint, so it flags output from ChatGPT, Claude, Gemini, Copilot and open-weight models alike.
Does it work on documents that were AI-assisted rather than AI-written?: Yes. Sentence-level scoring shows mixed documents as a patchwork — high-confidence AI passages sitting alongside clearly human ones — which is usually the signal you actually need.
Is the detector free?: Your first check is free with no signup. Heavier use is covered by the £9.99 24-hour Day Pass or the £39.99/month Professional plan.
What happens to the documents I upload?: Every uploaded file is permanently deleted from DocumePro servers within 60 minutes under our Non-Persistence Guarantee. Nothing is retained for analytics, shared with third parties, or used to train AI models.

Score a CV against a job spec · Run a metadata report · All forensic tools