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
Undetectable AIundetectable.ai
Closed source
Shared score components
Overall score12.61
Panel bypass score0.5%
Meaning preservation55.6%
Raw meaning cosine0.8856
Readability79.6%
Category consistency99.2%
Detector coverage231/231
Identical outputs0/33
Penalty points−18
Per-detector human-likelihood scores · measured outputs only
GPTZero0.0% 33/33
ZeroGPT0.5% 33/33
Copyleaks0.0% 33/33
Originality.ai3.4% 33/33
Pangram1.5% 33/33
Grammarly27.1% 33/33
QuillBot99.1% 33/33
Category scores · before entry-level penalties
Academic essay21.2
Application essay31.5
Blog post36.8
Business email36.1
Marketing copy33.4
Discussion board23.9
News article31.0
EvidenceAll 33 outputs