HumanizerEval

Leaderboard / Compare / WriteHuman vs Ryne

September 2026 review · same 49 texts

WriteHuman vs Ryne

In the September 2026 HumanizerEval review, WriteHuman and Ryne rewrote the same 49 English texts. WriteHuman cleared all 5 detectors on 32 of them; Ryne on 0. The difference is mostly Originality.ai: on 48 samples WriteHuman passed it where Ryne failed, and the reverse never happened.

Per-sample agreement

Per-sample agreement, September 2026. Counts are out of the 49 shared samples, computed at the 50% human-score threshold. Margin counts the samples only one tool passed, higher count first.
DetectorPassed more oftenMarginBoth passedNeither passed
Pangram 4.0WriteHuman24 to 2815
GPTZero 2026-09-13-baseWriteHuman45 to 040
Originality.ai turboWriteHuman48 to 001
ZeroGPT defaultWriteHuman19 to 0300
Winston AI 5.0WriteHuman36 to 0130
All 5 detectorsWriteHuman32 to 0017

Every count is computed from scores.csv at build time, not transcribed. “Neither passed” counts the texts that defeated both tools; on Pangram that was 15 of 49 here.

Quality and output discipline

Quality and output discipline, September 2026.
MeasureWriteHumanRyne
22%44%
49%80%
98%98%
65%94%
1.021.07
00
00
API defaultAggressive, beast mode
Websitewritehuman.ai ryne.ai

On per-sample quality scores, WriteHuman scored higher on 7 samples, Ryne on 32, and the two tied on 10. WriteHuman’s output was shorter on 27 of 49 samples; mean length ratios were 1.02 and 1.07, or 2% longer than the input and 7% longer than the input respectively.

Where each one outperforms the other

WriteHuman

  • Pangram: passed 32 of 49 against 10, including 24 samples where Ryne failed, against 2 the other way.
  • GPTZero: passed 49 of 49 against 4, including 45 samples where Ryne failed, against 0 the other way.
  • Originality.ai: passed 48 of 49 against 0, including 48 samples where Ryne failed, against 0 the other way.
  • ZeroGPT: passed 49 of 49 against 30, including 19 samples where Ryne failed, against 0 the other way.
  • Winston AI: passed 49 of 49 against 13, including 36 samples where Ryne failed, against 0 the other way.
  • Length: shorter output on 27 of 49 samples, which matters under a word limit.

Ryne

  • Pangram: passed 2 samples where WriteHuman failed, even though it was detected more often overall, passing 10 to 32.
  • Factual consistency: 44% against 22%.
  • Naturalness: 80% against 49%.
  • Stance preservation: 94% against 65%.

WriteHuman vs Ryne FAQ

Is WriteHuman better than Ryne?

In the September 2026 HumanizerEval review, WriteHuman cleared all 5 detectors on 32 of the 49 shared samples and Ryne on 0. Quality ran 7 samples to 32, with 10 ties.

Which humanizer is best for Pangram, WriteHuman or Ryne?

WriteHuman. In September 2026, WriteHuman passed Pangram on 32 of the 49 shared samples and Ryne on 10. On 24 samples WriteHuman passed where Ryne failed; the reverse happened on 2. Both failed on 15.

Which humanizer changes the text less?

Ryne. It scored higher on per-sample quality on 32 of 49 samples, against 7. WriteHuman’s outputs averaged 1.02× the input length and passed factual consistency on 22%; Ryne’s averaged 1.07× and passed on 44%. On stance preservation the split was 65% to 94%. WriteHuman produced the shorter output on 27 of 49 samples.

More comparisons