AI detector
Review AI signals in your text before you submit.
See probabilistic AI indicators at document and sentence level, with separate metrics and clear limits on what the result can — and cannot — tell you.
What the result means
AI detection is probabilistic, not proof of authorship.
AI detectors analyse patterns in writing and estimate how strongly those patterns resemble AI-generated text. They do not observe who wrote the document or how it was created.
False positives and false negatives are possible. Plagum presents the result as a signal to review, not as a final conclusion about authorship or misconduct.
- AI probability
- estimated likelihood that the document as a whole contains AI-generated writing
- Total AI rate
- share of sentences or fragments marked as likely AI-generated
- Sentence-level
- individual indicators help you see where stronger AI-associated patterns appear
The report
See where the AI signal comes from.
AI probability
Review the document-level probability instead of treating a single label as a final answer.
Total AI rate
See what share of sentences or fragments the system marks as likely AI-generated.
Sentence-level indicators
Inspect individual sentences to understand where stronger AI-associated patterns appear in the text.
How it works
From text to AI report in three steps.
-
01
Add your text.
Paste text, upload a PDF, DOCX, TXT, RTF or ODT file, or enter a URL. Each uploaded file can be up to 24 MB.
-
02
Analyse writing patterns.
Plagum evaluates statistical patterns associated with AI-generated writing and produces probabilistic indicators for the submitted text.
-
03
Review the report.
Review AI probability, Total AI rate and sentence-level indicators together with the surrounding text. The result should not be treated as proof of authorship.
Who it’s for
Built for people who want to review AI signals carefully.
Students
Review assignments, essays and papers before submission and see which passages may produce stronger AI indicators.
Researchers
Check academic drafts, articles and longer documents while keeping the result in context rather than treating it as a final conclusion.
Writers & editors
Review articles, drafts and client content when you need a clearer view of AI-associated writing patterns.
Use the result responsibly
An AI score should start a review, not end one.
AI detection can produce false positives and false negatives. A high score does not prove that a person used AI, and a low score does not prove that they did not.
Plagum should not be used as the sole basis for academic, disciplinary, employment or legal decisions. Human review and additional context remain necessary.
Submitted documents are not used to train AI models and are not automatically added to the general comparison database.
FAQ
Questions, answered plainly.
Still have a question? Contact support
