AI Humanizer BenchmarkGPTinf
AI humanizer · gptinf.com
In the October 2026 cycle, GPTinf placed #3 of 11 with an overall score of 80.74/100. AI Humanizer Benchmark tests GPTinf independently. The overall score weights bypass rate at 42%, meaning preservation at 32%, readability at 16% and consistency across categories at 10%, then subtracts any penalties.
- Last tested
- Sample size
- 33
- Methodology
- v1.1.0
Detector bypass rates
Average human-likelihood each detector gave this tool's outputs, shown as a percentage.
Performance by category
The overall-score formula applied within each category, before penalties.
Output quality
Meaning preservation is the rescaled embedding cosine similarity between input and output. Readability is a language model's rating of clarity, fluency and naturalness. Consistency shows how evenly bypass results hold across writing categories. The published data lists each value for every test.
No quality penalties this cycle. No meaning drift, length problems, refusals or unchanged outputs in any test.
Evidence: every GPTinf run
Each row is one October 2026 prompt that GPTinf rewrote, with a dot for each of the 7 detectors. Green means the detector judged the rewrite human. Red means it scored the rewrite as AI. An empty circle means no score came back.
12 of 33 outputs passed all 7 detectors. The other 21 got an AI verdict from at least one detector or are missing a score.
Each detector's result is converted to a human-likelihood score between 0 and 1. A score of 0 means the detector is sure the text is AI. A score of 1 means it is sure the text is human. Dots are green at 0.50 or above and red below. Hover a dot to see its exact score. The 0.50 cut-off only sets the dot colors. A tool's bypass rate on the leaderboard takes the median of the 7 detector scores for each test, then averages those medians. That gives a continuous number, so it will not match the share of green dots. The same goes for each per-detector rate. It is the average of that detector's scores (the number detector pages rank by), not a count of green dots.
Score history
Overall score in every published cycle that included this tool.
Access
- Starting price
- $9.99/mo
- Free tier
- Yes
- Public API
- Yes (required for inclusion)
Lowest paid plan at the monthly rate, without any annual discount. Price does not affect the score.
How we tested
The tool rewrote 33 texts from 7 writing categories on its default settings. Then 7 independent AI detectors scored each output. Anyone can recompute every score from the published data.
Compare with
Featured in
Embed this badge
Show GPTinf’s overall rank on your own site. The badge always uses the latest published cycle, so embedded copies update when the rank changes. Load the image from our URL and link it back to this page.
