Leaderboard / Humanizers / Grubby

September 2026 review · rank 3 of 10

Grubby AI detector test results

In the September 2026 HumanizerEval review, Grubby ranked 3 of 10 with a mean strict-bypass rate of 85.6%, behind StealthGPT Super at 96.8%. It cleared all 5 detectors on 18 of 50 texts. Review our full results for the best AI humanizer on the leaderboard page.

Grubby results, September 2026

Website

grubby.ai

Rank

3 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.

85.6%

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

18 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.

18 of 50 (36%), mean 39.8%, median 10.69%

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.

47 of 50 (94%), mean 93.6%, 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.

49 of 50 (98%), mean 98.2%, median 100%

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 96.0%, 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.

50 of 50 (100%), mean 99.3%, median 99.85%

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

72%

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

46%

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

98%

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

94%

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

1.11

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

0

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.

Academic

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

“Our bypass ai tool helps you achieve 99%+ human scores on even the most advanced AI detectors, including GPTZero, Originality.ai, ZeroGPT, Turnitin, Winston AI, Content at Scale, and Copyleaks.” (grubby.ai , retrieved October 1, 2026)

Measured

Strict bypass in September 2026: Pangram 36%, GPTZero 94%, Originality.ai 98%, ZeroGPT 100%, Winston AI 100%.

Verdict

Not verified. In September 2026, strict bypass was GPTZero 94% against the 99% claimed, Originality.ai 98% against the 99% claimed, ZeroGPT 100% against the 99% claimed and Winston AI 100% against the 99% claimed. Turnitin, Content at Scale and Copyleaks 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.

• Grubby vs StealthGPT

  Grubby cleared all 5 detectors on 1 sample where StealthGPT did not; the reverse happened on 29. Both cleared 17.

• Grubby vs WriteHuman

  Grubby cleared all 5 detectors on 2 samples where WriteHuman did not; the reverse happened on 17. Both cleared 15.

• Grubby vs Rephrasy

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

• Grubby vs Undetectable.ai

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

• Grubby vs RewriteAI

  Grubby cleared all 5 detectors on 10 samples where RewriteAI did not; the reverse happened on 5. Both cleared 8.

• Grubby vs HIX Bypass

  Grubby cleared all 5 detectors on 17 samples where HIX Bypass did not; the reverse never happened. Both cleared 1.

• Grubby vs Humbot

  Grubby cleared all 5 detectors on 18 samples where Humbot did not; the reverse never happened. Both cleared 0.

• Grubby vs BypassGPT

  Grubby cleared all 5 detectors on 18 samples where BypassGPT did not; the reverse never happened. Both cleared 0.

• Grubby vs Ryne

  Grubby cleared all 5 detectors on 18 samples where Ryne did not; the reverse never happened. Both cleared 0.