AI Humanizer BenchmarkStealthGPT
AI humanizer · stealthgpt.ai
StealthGPT is an AI humanizer independently benchmarked by AI Humanizer Benchmark. It ranked #7 of 11 in the September 2026 cycle with an overall score of 71.90/100, our composite measure of how reliably it evades AI detectors while preserving the original meaning and readability. Quality penalties deducted 3.0 points this cycle.
- Last tested
- 2026-09-03T16:11:14.775Z
- Sample size
- 33
- Methodology
- v1.0.0
Detector bypass rates
Percentage of samples classified as human-written by each detector.
Performance by category
Score across writing contexts.
Evidence: every run
Every prompt StealthGPT was run on in the September 2026 cycle, and how each of the 7 detectors scored the output. Green: the detector called it human-written. Red: it was caught.
8 of 33 outputs passed all 7 detectors; 25 were caught by at least one.
Each detector returns a human-likelihood on a common 0 to 1 scale, where 1 means it judged the text human-written and 0 means it flagged it as AI. A verdict counts as passed when that score is at least 0.50, the midpoint of the scale; hover any dot for the exact value. That threshold exists only to draw the dots: the bypass rate on the leaderboard is the mean of each test's median score across the 7 detectors, a continuous number, so a tool's pass count and its bypass rate will not be the same figure. Likewise a detector-rate row is the mean of that detector's scores, the same number the detector pages rank by, not a count of green dots.
Access
- Free tier
- Yes
- Public API
- Yes (required for inclusion)
How we tested
33 prompts across 7 writing categories, run through the tool on its default settings and scored by 7 independent AI detectors. Every score is reproducible from the published data.