Leaderboard / Humanizers / Undetectable.ai

September 2026 review · rank 5 of 10

Undetectable.ai AI detector test results

In the September 2026 HumanizerEval review, Undetectable.ai ranked 5 of 10 with a mean strict-bypass rate of 84.0%, behind StealthGPT Super at 96.8%. It cleared all 5 detectors on 15 of 50 texts. Review our full results for the best AI humanizer on the leaderboard page.

Undetectable.ai results, September 2026

Website

undetectable.ai

Rank

5 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.

15 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.

16 of 50 (32%), mean 31.4%, median 8.21%

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.

48 of 50 (96%), mean 94.7%, 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.

47 of 50 (94%), mean 93.1%, median 99.99%

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 95.0%, 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.

49 of 50 (98%), mean 97.5%, median 99.84%

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

26%

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

28%

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

94%

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

78%

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

1.21

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

7

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.

University / General Writing / More Human, v11sr

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

“A 99% bypass rate.” (undetectable.ai , retrieved September 18, 2026)

Measured

Strict bypass in September 2026: Pangram 32%, GPTZero 96%, Originality.ai 94%, ZeroGPT 100%, Winston AI 98%.

Verdict

Not verified. Mean strict bypass was 84.0% in September 2026, against the 99% claimed.

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.

• Undetectable.ai vs StealthGPT

  Undetectable.ai cleared all 5 detectors on 0 samples where StealthGPT did not; the reverse happened on 31. Both cleared 15.

• Undetectable.ai vs WriteHuman

  Undetectable.ai cleared all 5 detectors on 0 samples where WriteHuman did not; the reverse happened on 18. Both cleared 14.

• Undetectable.ai vs Grubby

  Undetectable.ai cleared all 5 detectors on 6 samples where Grubby did not; the reverse happened on 9. Both cleared 9.

• Undetectable.ai vs Rephrasy

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

• Undetectable.ai vs RewriteAI

  Undetectable.ai cleared all 5 detectors on 8 samples where RewriteAI did not; the reverse happened on 6. Both cleared 7.

• Undetectable.ai vs HIX Bypass

  Undetectable.ai cleared all 5 detectors on 14 samples where HIX Bypass did not; the reverse never happened. Both cleared 1.

• Undetectable.ai vs Humbot

  Undetectable.ai cleared all 5 detectors on 15 samples where Humbot did not; the reverse never happened. Both cleared 0.

• Undetectable.ai vs BypassGPT

  Undetectable.ai cleared all 5 detectors on 15 samples where BypassGPT did not; the reverse never happened. Both cleared 0.

• Undetectable.ai vs Ryne

  Undetectable.ai cleared all 5 detectors on 15 samples where Ryne did not; the reverse never happened. Both cleared 0.