November 2026: 15 services · five detectors · updated scoring

November 2026 · same 50 inputs

Grubby vs Solvely

November 2026 · methodology v2 · provisional automated quality. 15 humanizers, 50 identical inputs each, five detectors and eight quality metrics. Read the scoring rules and post-run changes.
November 2026: overall = 70% weighted bypass + 30% quality. All scores out of 100.
Overall rankHumanizerOverall averageOverall medianWeighted bypassQuality
9GrubbyModel: Not exposed · Mode: Academic28.3626.8510.40%70.281.19
12SolvelyModel: Not exposed · Mode: Humanize; Original result25.9725.706.10%72.321.15

Grubby: detector breakdown

Strict passes on 50 outputs; raw human-score averages are shown separately and have different meanings across vendors.
DetectorModelWeightPassesPass rateNative human-score average
Pangram 4.14.140%1/502.00%8.68
GPTZero 5.0b5.0b / 2026-11-04-base30%4/508.00%6.87
Originality.ai 5.0model-5.020%0/500.00%7.42
Winston AI 5.25.25%50/50100.00%99.79
ZeroGPTVendor-managed; no public version identifier5%22/5044.00%70.99

Solvely: detector breakdown

Strict passes on 50 outputs; raw human-score averages are shown separately and have different meanings across vendors.
DetectorModelWeightPassesPass rateNative human-score average
Pangram 4.14.140%1/502.00%7.09
GPTZero 5.0b5.0b / 2026-11-04-base30%0/500.00%0.00
Originality.ai 5.0model-5.020%0/500.00%3.98
Winston AI 5.25.25%44/5088.00%86.95
ZeroGPTVendor-managed; no public version identifier5%9/5018.00%38.09

Overall weights humanization 70% and quality 30%. Rankings use automated quality scores with unresolved evidence flags; neither a detector pass nor a small score gap proves a universal advantage.

Per-output scores CSV · Frozen inputs and prompts CSV · Scoring manifest · Verification bundle · GitHub report · Earlier editions