Leaderboard / Humanizers / Rephrasy

September 2026 review · rank 4 of 10

Rephrasy AI detector test results

In the September 2026 HumanizerEval review, Rephrasy ranked 4 of 10 with a mean strict-bypass rate of 84.0%, behind StealthGPT Super at 96.8%. It cleared all 5 detectors on 20 of 50 texts. Review our full results for the best AI humanizer on the leaderboard page.

Rephrasy results, September 2026

Website

rephrasy.ai

Rank

4 of 10

Mean strict bypassThe average of the five detector pass rates: the share of texts each detector called human-written. The leaderboard is ranked by this number.

84.0%

Passed all 5 detectorsTexts that all five detectors called human-written.

20 of 50

Pangram 4.0 strict bypassHow many texts Pangram called human-written (a human score above 50%), with the mean and median human score it gave across all texts.

23 of 50 (46%), mean 47.2%, median 34.53%

GPTZero 2026-09-13-base strict bypassHow many texts GPTZero called human-written (a human score above 50%), with the mean and median human score it gave across all texts.

44 of 50 (88%), mean 87.3%, median 100%

Originality.ai turbo strict bypassHow many texts Originality.ai called human-written (a human score above 50%), with the mean and median human score it gave across all texts.

43 of 50 (86%), mean 85.1%, median 99.93%

ZeroGPT default strict bypassHow many texts ZeroGPT called human-written (a human score above 50%), with the mean and median human score it gave across all texts.

50 of 50 (100%), mean 92.7%, median 100%

Winston AI 5.0 strict bypassHow many texts Winston AI called human-written (a human score above 50%), with the mean and median human score it gave across all texts.

50 of 50 (100%), mean 98.8%, median 99.78%

Factual consistency passShare of outputs where key claims, numbers, citations, and named entities remain accurate and correctly attributed.

36%

Naturalness passShare of outputs where wording is fluent and free of awkward or clunky phrasing.

30%

Syntax integrity passShare of outputs where sentences are grammatically complete and free of disruptive run-ons or punctuation errors.

92%

Stance preservation passShare of outputs where the source's position, tone, and intent are not reversed or distorted.

60%

Mean length ratioOutput word count divided by input word count, averaged across texts. 1.00 is the same length; 1.10 is 10% longer.

1.13

Samples over 1.4× inputOutputs more than 1.4× the input’s word count.

3

Samples under 0.6× inputOutputs shorter than 0.6× the input’s word count.

0

Settings usedThe mode each tool ran on: its strongest advertised setting on the cheapest paid tier.

v4

Recompute any of this from scores.csv. Strict bypass means an overall human verdict on a human score above 50%.

What the vendor claims, and what we measured

Claim

“We guarantee 100% pass rates on every major detector.” (rephrasy.ai , retrieved October 1, 2026) The same answer says Rephrasy bypasses all major AI detectors including Turnitin, GPTZero, Originality.ai, Copyleaks, ZeroGPT, Sapling, and Content at Scale.

Measured

Strict bypass in September 2026: Pangram 46%, GPTZero 88%, Originality.ai 86%, ZeroGPT 100%, Winston AI 100%.

Verdict

Not verified. In September 2026, strict bypass was GPTZero 88% against the 100% claimed, Originality.ai 86% against the 100% claimed, ZeroGPT 100% against the 100% claimed, Winston AI 100% against the 100% claimed and Pangram 46% against the 100% claimed. Turnitin, Copyleaks, Sapling and Content at Scale are not in this panel.

Review history

Head to head

Because every tool saw the same texts, these pages report per-sample agreement, how often each tool bypassed detection on the same input, rather than a difference of averages.

• Rephrasy vs StealthGPT

  Rephrasy cleared all 5 detectors on 0 samples where StealthGPT did not; the reverse happened on 26. Both cleared 20.

• Rephrasy vs WriteHuman

  Rephrasy cleared all 5 detectors on 1 sample where WriteHuman did not; the reverse happened on 14. Both cleared 18.

• Rephrasy vs Grubby

  Rephrasy cleared all 5 detectors on 9 samples where Grubby did not; the reverse happened on 7. Both cleared 11.

• Rephrasy vs Undetectable.ai

  Rephrasy cleared all 5 detectors on 9 samples where Undetectable.ai did not; the reverse happened on 4. Both cleared 11.

• Rephrasy vs RewriteAI

  Rephrasy cleared all 5 detectors on 11 samples where RewriteAI did not; the reverse happened on 4. Both cleared 9.

• Rephrasy vs HIX Bypass

  Rephrasy cleared all 5 detectors on 19 samples where HIX Bypass did not; the reverse never happened. Both cleared 1.

• Rephrasy vs Humbot

  Rephrasy cleared all 5 detectors on 20 samples where Humbot did not; the reverse never happened. Both cleared 0.

• Rephrasy vs BypassGPT

  Rephrasy cleared all 5 detectors on 20 samples where BypassGPT did not; the reverse never happened. Both cleared 0.

• Rephrasy vs Ryne

  Rephrasy cleared all 5 detectors on 20 samples where Ryne did not; the reverse never happened. Both cleared 0.