November 2026 · Independent humanizer benchmarks. Explore the evidence

Compare AI humanizers

Pick two to four entries and inspect them side by side. All numbers come from the same shared source set and measurement profiles. An asterisk marks detector-dependent, provisional values.

Pick 2 to 4 entries 1 selected

November 2026
UndetectedGPTundetectedgpt.ai
Closed source
Shared score components
Overall score50.65
Panel bypass score61.3%
Meaning preservation46.8%
Raw meaning cosine0.8661
Readability79.8%
Category consistency69.3%
Detector coverage231/231
Identical outputs0/33
Penalty points−10
Per-detector human-likelihood scores · measured outputs only
GPTZero9.4% 33/33
ZeroGPT29.1% 33/33
Copyleaks75.8% 33/33
Originality.ai8.2% 33/33
Pangram60.8% 33/33
Grammarly93.1% 33/33
QuillBot99.9% 33/33
Category scores · before entry-level penalties
Academic essay69.4
Application essay45.6
Blog post57.1
Business email37.7
Marketing copy70.1
Discussion board69.0
News article76.1
EvidenceAll 33 outputs