Winston AI gave Superhumanizer its strongest detector result, with 31 passes out of 50. StealthGPT Super passed 48 Winston checks and led on the other four detectors. Super also scored 78.00 for factual consistency, compared with 26.00 for Superhumanizer. For publishing workflows, compare figures, names and claims against your original draft before shipping a rewrite.
Read the full November benchmark and Superhumanizer updateStealthGPT: Superhumanizer Comparison and Alternative
Compare the tested humanizers on the same drafts, then choose the writing workflow you need.
StealthGPT Super passed 44 of 50 Pangram checks, compared with 1 for Superhumanizer. Super also scored higher on writing quality, 73.48 versus 51.88, and placed second overall to Superhumanizer’s fifteenth.
- 16 humanizers tested
- 50 matched inputs per tool
- Five detectors
HumanizerEval · November 2026
StealthGPT Super vs Superhumanizer on 50 matched inputs
The November benchmark compares 16 humanizers on the same 50 English inputs. Superhumanizer joined on November 18 and the original results were carried forward. Detector rows show passes. Writing quality scores are automated and provisional.
StealthGPT Super
Super, Quality, rephrase enabled, API
Superhumanizer
Super Ultra engine, Transform mode, Starter API plan
| Measure | StealthGPT Super | Superhumanizer |
|---|---|---|
| Overall score | 70.34/100 (#2) | 19.41/100 (#15) |
| Weighted detector pass rate | 69.00% | 5.50% |
| Writing quality | 73.48/100 (#3) | 51.88/100 (#14) |
| Pangram 4.1 | 44/50 (88%) | 1/50 (2%) |
| GPTZero 5.0b | 30/50 (60%) | 2/50 (4%) |
| Originality.ai 5.0 | 19/50 (38%) | 0/50 (0%) |
| Winston AI 5.2 | 48/50 (96%) | 31/50 (62%) |
| ZeroGPT | 34/50 (68%) | 4/50 (8%) |
| Readability | 68.80/100 | 47.20/100 |
| Meaning preservation | 67.20/100 | 48.40/100 |
| Factual consistency | 78.00/100 | 26.00/100 |
The overall score combines 70% weighted detector passes and 30% writing quality. Detector weights are 40% Pangram, 30% GPTZero, 20% Originality.ai, 5% Winston AI and 5% ZeroGPT. Winston uses a document Human Score of at least 80. The other detectors use qualifying human labels. HumanizerEval is operated by the StealthGPT team.
What the comparison shows
Frequently Asked Questions
Still have more questions?
Learn more in our support page.
No on most tested drafts. Superhumanizer passed 1 of 50 checks (2%). HumanizerEval’s November 2026 benchmark used Pangram 4.1. StealthGPT Super passed 44 of the same 50 drafts. See the dated detector results.
No on most tested drafts. Superhumanizer passed 2 of 50 checks (4%). HumanizerEval’s November 2026 benchmark used GPTZero 5.0b. StealthGPT Super passed 30 of the same 50 drafts. See the dated detector results.
No. Superhumanizer passed 0 of 50 checks. HumanizerEval’s November 2026 benchmark used Originality.ai 5.0. StealthGPT Super passed 19 of the same 50 drafts. See the dated detector results.
No on most tested drafts. Superhumanizer passed 4 of 50 checks (8%). HumanizerEval’s November 2026 benchmark used ZeroGPT. StealthGPT Super passed 34 of the same 50 drafts. See the dated detector results.
Yes on 31 of 50 tested drafts (62%). HumanizerEval’s November 2026 benchmark used Winston AI 5.2. StealthGPT Super passed 48 of the same 50 drafts. See the dated detector results.
No. None of its 50 tested outputs passed all five detectors. The November 2026 panel used Pangram, GPTZero, Originality.ai, ZeroGPT and Winston AI. StealthGPT Super passed all five on 13 of the same 50 drafts. Compare the full benchmark.
HumanizerEval tested Superhumanizer’s provider-advertised Super Ultra API engine in Transform mode on the Starter API plan. The provider managed failover and did not expose the exact build that served the requests.
HumanizerEval calculates the overall score using 70% weighted detector passes and 30% provisional automated writing quality. The quality rubric includes readability, factual consistency and six other measures. The scores apply to the configurations shown.
See how StealthGPT fits your writing workflow
Explore the AI Humanizer and see how Super revises a sample draft.

