October 2026 Benchmark Results
Phrasly
Seventh overall at 63.1 despite ranking second for detector bypass at 90.5; meaning (56.8) and readability (54.3) lower its overall result.
Scorecard
By detector
How Phrasly performs against each detector we test.
| Detector | Score | Pass rate |
|---|---|---|
| Pangram | 88.7 | 89% |
| GPTZero | 91.9 | 92% |
| Originality.ai | 91.4 | 91% |
| Copyleaks | 90.2 | 90% |
| ZeroGPT | 90.5 | 91% |
At a glance
Phrasly sits 2.8 points behind HIX Bypass and 2.7 points ahead of SmartHumanizer in the October overall ranking. Its 90.5 detector-bypass score is second only to GPTHuman.
Where it stands out
Phrasly places second on all five detector-specific rankings. It scores 88.7 against Pangram, 91.9 against GPTZero, 91.4 against Originality.ai, 90.2 against Copyleaks and 90.5 against ZeroGPT.
The trade-off
Meaning preservation at 56.8 and readability at 54.3 both rank tenth in the October field. That large gap explains why excellent detector performance does not translate into a top overall position.
Bottom line
Phrasly is the strongest detector specialist after GPTHuman in the October data, but it carries the clearest quality trade-off. Any output should be reviewed closely for preserved meaning and readable phrasing.
October 2026 results
The benchmark case for Phrasly
The current cycle places Phrasly second for detector bypass but seventh overall because its published quality scores are much lower.
Overall: 63.1
Ranked #7 of 10 tools, 2.7 points ahead of eighth place.
Detector bypass: 90.5
Ranked #2 among the 10 October tools for detector bypass.
Meaning: 56.8
Ranked #10 among the 10 October tools for meaning preservation.
Readability: 54.3
Ranked #10 among the 10 October tools for readability.
Related research
Explore Humanizer Arena research
See separate experiments and the site methodology for additional context.