Free AI slop detector.
See every tell.

The server preserves each sentence’s exact position, sends bounded batches to TypeSafe’s Jev model, and asks for both a 0–4 style score and the strongest matching signal from a Wikipedia-derived rubric. The stopwatch covers the full request-to-render time.

Only sentences that cross the detector’s evidence threshold. Use the arrows to move through detections and see the original sentence number, signal category, Jev score, confidence, and matched wording.

No. It marks a recognizable writing pattern, not authorship. Human prose can contain the same patterns, so every result needs human judgment and should never be the sole basis for a high-stakes decision.

No. AI detectors look at more than common, obvious tells, and removing a few highlighted phrases cannot guarantee an undetectable result. For a complete rewrite designed to reduce detection signals, visit Make your content undetectable with StealthGPT ↗

Yes. Revise any detected sentence directly in its review card, then choose “Apply & recheck” to ask Jev whether the signal remains. The revision is applied only after a successful check, and every other highlight stays in place. You can still choose “Edit in place” to change the full document.

Up to 10,000 words at once, with no sentence-count rule. Extremely sentence-dense documents are grouped into adjacent passages so the scan remains reliable and bounded.

This app has no accounts or document database and does not save pasted text in application code. Text is processed by Vercel and TypeSafe to provide the scan, so confidential material should not be submitted. See the privacy page for details.