Why a detector can still flag a result

A detector score is a statistical estimate, not proof of who wrote a text. Different detectors use different models, so the same text can score differently on each, and their models change over time.

What you can do

  • Check the sentence-by-sentence breakdown on your result and edit the flagged sentences in your own words.
  • Run longer passages rather than short fragments. Very short texts are hard for any detector to judge.
  • Avoid pasting AI phrasing back into the result after the rewrite.

We do not guarantee any score on any detector. If a completed rewrite scores below 90 on ours, the words are returned automatically once per document. If you think a result is wrong, tell us and include the document.

Still need help?

Send us a message and we will answer, usually within a day.

Contact support