Skip to content
The September 2026 results are in: WriteHuman holds #1, and newcomer SupWriter debuts at #5.Read the full analysis →
HumanizerBench

Head to head · September 2026 cycle

Penlify vs StealthGPT

StealthGPT finished 28.5 points ahead of Penlify in the September 2026 cycle, 65.10 to 36.57, ranking #9 against #14 in a field of 14. StealthGPT scored higher on 5 of 5 detectors and 7 of 7 writing categories. Penlify came out ahead on consistency across categories (100.0 vs 63.1), meaning preservation (62.7 vs 59.4) and readability (71.8 vs 70.9).

36.57 /100
overall score
#14 /14
rank -
+28.53
StealthGPT leads →
Detectors
0 5
Categories
0 7
Components
3 1 (1 tie)
65.10 /100
overall score
#9 /14
rank ▲ 3
Last tested
Prompts
33
Methodology
v1.2.0

Score components

What the overall score is made of: bypass rate weighs 42%, meaning preservation 32%, readability 16%, consistency across categories 10%. Output-quality penalties come off the total. How scoring works →

Bypass rate StealthGPT by 79.6
0.0 79.6
Meaning preservation CIs overlap Penlify by 3.3
62.7 59.4
Readability Penlify by 0.9
71.8 70.9
Consistency Penlify by 36.9
100.0 63.1
Penalties fewer is better Even
−5.0 −5.0

CIs overlap marks a gap that sits inside both tools' published 95% confidence intervals for that component; it may not survive another cycle.

Detector by detector

Share of each tool's outputs that a detector classified as human-written. StealthGPT took 5 of 5.

GPTZero StealthGPT by 77.4
0.0 77.4
Originality.ai StealthGPT by 89.5
0.0 89.5
Copyleaks StealthGPT by 81.8
0.0 81.8
Winston AI StealthGPT by 66.2
0.4 66.6
ZeroGPT StealthGPT by 25.9
43.0 68.9

Writing categories

Category score per writing context: bypass credit only for real rewrites, blended with the category's own meaning preservation and readability. Small categories swing; the prompt count is shown on each row. How category scores work →

Academic Essay Application Essay Blog Post Business Email Marketing Copy Discussion Board News Article Academic Essay: Penlify 27.0 · StealthGPT 79.4 Application Essay: Penlify 35.9 · StealthGPT 60.3 Blog Post: Penlify 37.3 · StealthGPT 60.7 Business Email: Penlify 39.9 · StealthGPT 42.7 Marketing Copy: Penlify 38.8 · StealthGPT 73.1 Discussion Board: Penlify 32.1 · StealthGPT 60.1 News Article: Penlify 36.6 · StealthGPT 64.9
Penlify StealthGPT Outer ring = 100
Academic Essay 6 prompts StealthGPT by 52.4
27.0 79.4
Application Essay 6 prompts StealthGPT by 24.4
35.9 60.3
Blog Post 6 prompts StealthGPT by 23.4
37.3 60.7
Business Email 3 prompts StealthGPT by 2.7
39.9 42.7
Marketing Copy 6 prompts StealthGPT by 34.3
38.8 73.1
Discussion Board 3 prompts StealthGPT by 28.0
32.1 60.1
News Article 3 prompts StealthGPT by 28.3
36.6 64.9

Where each one wins

Every measure above that one tool won outright, largest margin first within each group.

Penlify

3 wins
  • Consistency Score 100.0 vs 63.1
  • Meaning preservation Score 62.7 vs 59.4
  • Readability Score 71.8 vs 70.9

StealthGPT

13 wins
  • Bypass rate Score 79.6 vs 0.0
  • Originality.ai Detector 89.5 vs 0.0
  • Copyleaks Detector 81.8 vs 0.0
  • GPTZero Detector 77.4 vs 0.0
  • Winston AI Detector 66.6 vs 0.4
  • ZeroGPT Detector 68.9 vs 43.0
  • Academic Essay Category 79.4 vs 27.0
  • Marketing Copy Category 73.1 vs 38.8
  • News Article Category 64.9 vs 36.6
  • Discussion Board Category 60.1 vs 32.1
  • and 3 more

Questions people ask

Is Penlify better than StealthGPT?

StealthGPT finished 28.5 points ahead of Penlify in the September 2026 cycle, 65.10 to 36.57, ranking #9 against #14 in a field of 14. StealthGPT scored higher on 5 of 5 detectors and 7 of 7 writing categories. Both tools ran the same 33 prompts, scored by the same 5 detectors, under methodology v1.2.0; every input, output and verdict is published.

Which bypasses AI detectors better, Penlify or StealthGPT?

StealthGPT posted the higher bypass rate, 79.6 vs 0.0, and scored higher on 5 of 5 detectors. StealthGPT led on GPTZero, Originality.ai, Copyleaks, Winston AI and ZeroGPT.

Do Penlify and StealthGPT keep the original meaning?

Penlify preserved meaning better, 62.7 vs 59.4 on our 0–100 similarity scale, though the confidence intervals overlap. On readability, Penlify rated higher, 71.8 vs 70.9. Penalties this cycle: Penlify −5.0, StealthGPT −5.0.