Rephrasy vs. UndetectedGPT
How do the two commercial entries compare on readability, given their different overall rankings?
Across 33 matched rewrites with numeric readability scores, Rephrasy scored higher on 7, UndetectedGPT on 23, and 3 tied. Their November composite scores were 55.34 and 50.65 out of 100, respectively.
Results on the same source texts
Each row compares both entries on the same source and includes pairs with two numeric scores. Meaning uses recorded embedding cosine. Detector scores reflect reported human-likelihood.
| Metric | Rephrasy higher | UndetectedGPT higher | Tied | Measured pairs |
|---|---|---|---|---|
| Readability | 7 | 23 | 3 | 33 / 33 |
| Meaning preservation | 10 | 23 | 0 | 33 / 33 |
| GPTZero | 20 | 3 | 10 | 33 / 33 |
| ZeroGPT | 17 | 6 | 10 | 33 / 33 |
| Copyleaks | 5 | 1 | 27 | 33 / 33 |
| Originality.ai | 29 | 4 | 0 | 33 / 33 |
| Pangram | 17 | 10 | 6 | 33 / 33 |
| Grammarly | 7 | 11 | 15 | 33 / 33 |
| QuillBot | 5 | 9 | 19 | 33 / 33 |
Read the matched rewrites
These examples show the largest measured gaps in readability, with one example per writing category.
- Blog post ? S16
Rephrasy rewrite (46.7%) ? UndetectedGPT rewrite (85.0%) - Academic essay ? S23
Rephrasy rewrite (43.3%) ? UndetectedGPT rewrite (80.0%) - Application essay ? S04
Rephrasy rewrite (63.3%) ? UndetectedGPT rewrite (76.7%)
Scope and source
Both entries use historical published commercial rewrites. The comparison uses the recorded outputs in the November 2026 dataset, version 2026-11-v1. See the methodology for metric definitions and limitations.
Model profiles: Rephrasy and UndetectedGPT.