StealthGPT ranked #2 of 16 in the November 2026 cycle with an overall score of 69.47/100, after losing 1 point to penalties for length deflation. Copyleaks passed its output most often (90.0%), and Originality.ai passed it least often (28.8%). It did best on blog posts and worst on marketing copy.
StealthGPT, from XYZ AI, Inc., combines an AI humanizer with an essay writer, an SEO writer and an AI detector, and has an iOS app and a Chrome extension. The humanizer offers Stealth Lite and Super rewrite modes plus tone settings from high school to Ph.D. level. No ongoing free humanizer plan is confirmed; a paid introductory trial is available.
Watch the 2-minute review of StealthGPT
- Overall score
- 69.47/100
- Detector bypass
- 66.0%
- Runner-up
- #2of 16
- Prompts tested
- 30
Detector Bypass Rates
Each detector's average human-likelihood score on this tool's outputs, where 100 means every output was scored as human.
Performance by Category
Category score across writing contexts: bypass credit only for real rewrites, blended with each category's meaning and readability. How it's computed
Strengths
- Readability of 77.7, the highest this cycle (field median 59.5)
- Copyleaks classified 90.0% of its runs as human-written
- Best writing category: blog posts (75.4)
Weaknesses
- Meaning preservation of 67.3, below the field median of 72.8
- Lost 1 point to penalties for length deflation
- Originality.ai classified only 28.8% of its runs as human-written
From the November 2026 cycle, compared with the other 15 tools ranked in that cycle.
Evidence: every run
Every prompt StealthGPT was run on in November 2026, and how each of the 6 detectors scored the output. Green: the detector called it human-written. Red: it was caught.
3 of 30 outputs passed all 6 detectors; 27 were caught by at least one.
| Prompt | GPT GPTZero | Pan Pangram | Win Winston AI | Zero ZeroGPT | Copy Copyleaks | Orig Originality.ai | Mean | Recording |
|---|---|---|---|---|---|---|---|---|
| Landing Copy · claude-sonnet-5-5 | caught | caught | passed | passed | passed | caught | 0.51 | |
| Listicle Blog · claude-sonnet-5-5 | passed | passed | passed | passed | passed | caught | 0.78 | |
| Personal Statement · gemini-3-8-flash | caught | caught | passed | passed | passed | caught | 0.46 | |
| News Article · gpt-6-sol | passed | caught | passed | passed | passed | caught | 0.55 | |
| Cover Letter · gpt-6-sol | caught | caught | caught | caught | caught | caught | 0.01 | |
| Lit Review · gpt-6-sol | passed | passed | passed | passed | passed | passed | 0.95 | |
| Howto Blog · claude-sonnet-5-5 | passed | passed | passed | caught | passed | caught | 0.72 | |
| Personal Statement · gpt-6-sol | passed | passed | passed | passed | passed | caught | 0.76 | |
| Product Desc · claude-sonnet-5-5 | passed | caught | caught | passed | passed | caught | 0.49 | |
| Landing Copy · gemini-3-8-flash | passed | caught | passed | passed | passed | caught | 0.62 | |
| Argumentative Essay · claude-sonnet-5-5 | passed | passed | passed | passed | passed | passed | 0.97 | |
| News Article · claude-sonnet-5-5 | caught | passed | passed | passed | passed | caught | 0.61 | |
| Landing Copy · gpt-6-sol | passed | passed | passed | passed | passed | caught | 0.91 | |
| Listicle Blog · gpt-6-sol | passed | passed | passed | passed | passed | caught | 0.80 | |
| Product Desc · gemini-3-8-flash | passed | passed | passed | caught | passed | passed | 0.86 | |
| Cover Letter · gemini-3-8-flash | caught | caught | caught | caught | caught | caught | 0.04 | |
| Argumentative Essay · gpt-6-sol | passed | passed | passed | passed | passed | caught | 0.78 | |
| Product Desc · gpt-6-sol | passed | passed | caught | caught | passed | caught | 0.51 | |
| Discussion Post · gpt-6-sol | caught | passed | passed | passed | passed | caught | 0.55 | |
| Discussion Post · claude-sonnet-5-5 | passed | caught | passed | passed | passed | caught | 0.66 | |
| Lit Review · gemini-3-8-flash | caught | caught | caught | caught | caught | caught | 0.01 | |
| Howto Blog · gpt-6-sol | passed | passed | passed | caught | passed | passed | 0.84 | |
| Argumentative Essay · gemini-3-8-flash | passed | passed | passed | passed | passed | caught | 0.80 | |
| Personal Statement · claude-sonnet-5-5 | passed | passed | passed | passed | passed | caught | 0.73 | |
| News Article · gemini-3-8-flash | passed | caught | passed | passed | passed | caught | 0.68 | |
| Discussion Post · gemini-3-8-flash | passed | passed | passed | passed | passed | passed | 0.99 | |
| Lit Review · claude-sonnet-5-5 | passed | passed | passed | passed | passed | caught | 0.81 | |
| Cover Letter · claude-sonnet-5-5 | passed | passed | passed | passed | passed | caught | 0.74 | |
| Listicle Blog · gemini-3-8-flash | passed | caught | passed | passed | passed | passed | 0.77 | |
| Howto Blog · gemini-3-8-flash | passed | passed | passed | caught | passed | passed | 0.88 | |
| Detector rate (mean human-likelihood) | 71.4 | 60.9 | 81.1 | 63.6 | 90.0 | 28.8 hardest | ||
| Runs passed | 23/30 | 19/30 | 25/30 | 22/30 | 27/30 | 7/30 |
A dot is green when the detector's human-likelihood score for that run was 0.50 or higher. The detector rate is the mean of that detector's scores across the runs, the same number the detector pages rank by; it is not the count of green dots, which the Runs passed row shows. Hover a dot for the score.
How to read these scores
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. On these pages a verdict counts as passed when that score is at least 0.50, the midpoint of the detector's own scale. That threshold exists only to draw the chips: the bypass rate on the leaderboard is the mean of each test's scores from the 6 detectors, a continuous number, so a tool's pass count and its bypass rate will not be the same figure.
History
Overall score compared with other ranked tools.
8.4 points above the median of 15 other tools in November 2026.
The shaded band is the 25th–75th percentile of other tools in each cycle. Scores are out of 100; higher is better.
Monthly pricing
Standard USD rates, billed month to month. Annual discounts and promotions excluded.
-
Pro
$29.99/mo50,000 words per month.
- Up to 1,000 input words per use.
-
Max
$49.99/mo200,000 words per month.
- Up to 2,000 input words per use.
-
Enterprise
$179.99/mo1,000,000+ words per month.
- 10,000+ input words per use, as listed on the pricing page.
Free access and trials
No ongoing free humanizer plan is confirmed. The current introductory offer is $1 for 3 days, followed by the regular monthly Pro or Max rate.
- Optional Samurai add-ons are excluded from the base prices. The finite monthly quotas do not specify input versus output word accounting.
Sample output
One AI-written passage from the November 2026 cycle and what StealthGPT returned for it. This prompt was picked by a fixed rule, not by how the tool scored.
AI-written input
Every tool on this prompt →Cover Letter · Application Essay · written by claude-sonnet-5-5 · 410 words
[Your Name] [Your Address] | [City, State ZIP] | [Phone] | [Email]
[Date]
[Hiring Manager's Name] [Department Name] City of [City Name] [Address]
Dear [Hiring Manager's Name],
I am writing to apply for the Policy Analyst position with the City of [City Name]. With a master's degree in public policy and four years of experience evaluating programs and advising local decision-makers, I am eager to help [City Name] craft practical, evidence-based policies that improve residents' daily lives.
In my current role at the [Organization Name], I analyze housing, transportation, and public safety initiatives for a region of 300,000 residents. Last year, I led an evaluation of a rental assistance pilot that revealed administrative barriers preventing nearly 20 percent of eligible households from enrolling. My recommendations, which included simplifying the application and partnering with community organizations for outreach, were adopted by the county and increased enrollment by 35 percent within six months. This experience taught me that strong policy analysis must pair rigorous data with a clear understanding of how programs work on the ground.
I am skilled in quantitative and qualitative research methods, including cost-benefit analysis, survey design, and statistical modeling in R and Excel. Equally important, I know how to translate complex findings into concise briefs and presentations that busy elected officials, department heads, and residents can use. I have presented to city council committees and facilitated public workshops, and I value the give-and-take of hearing community concerns and incorporating them into recommendations.
I am particularly drawn to [City Name]'s commitment to [specific initiative, e.g., expanding affordable housing, its climate action plan, or its Vision Zero goals]. I would welcome the chance to contribute to this work, whether by analyzing budget tradeoffs, researching best practices in peer cities, or evaluating outcomes to ensure programs deliver on their promises. Working in local government appeals to me because the results of good policy are visible and immediate: safer streets, stable housing, and responsive services.
Collaboration is central to my approach. I have worked alongside engineers, budget officers, legal staff, and nonprofit leaders, and I understand that effective policy depends on building trust across departments and with the public. I am organized, deadline-driven, and comfortable managing multiple projects while maintaining accuracy and transparency.
Thank you for considering my application. I would welcome the opportunity to discuss how my skills and commitment to public service can support the City of [City Name]. I can be reached at [phone] or [email].
Sincerely,
[Your Name]
Show the full input Show less
StealthGPT output
475 words (input 410)
[Your Name] [Your Address] | [City, State ZIP] | [Phone] | [Email]
[Date]
[Hiring Manager's Name] [Department Name] City of [City Name] [Address]
Dear [Hiring Manager's Name],
I’m excited to apply for the Policy Analyst position with the City of [City Name]. With a master’s degree in public policy and four years of experience conducting program evaluations and informing policymaking at the local level, I’d love the opportunity to conduct practical, evidence-informed policy and program analysis to help improve outcomes and quality of life for [City Name] residents.
In my current position at the [Organization Name], I conduct program and policy analysis on issues such as housing, transportation, and public safety to inform policymaking among local decision-makers. For a region of 300,000 residents, I most recently led an evaluation of a rental assistance pilot program to assess implementation challenges and areas for improvement. After identifying administrative barriers that may have prevented close to 20 percent of eligible households from enrolling, I made recommendations to the county for simplifying the application process and leveraging community-based organizations to increase awareness about the program. The county adopted many of these recommendations, and enrollment increased by 35 percent within six months. My career in policy analysis has underscored the importance of not only strong technical skills, but an understanding of how policies and programs operate on the ground to effectively inform decision-making.
I have strong quantitative and qualitative research skills, including cost-benefit analysis, survey design, and statistical analysis/modeling using R and Excel. I’m adept at communicating technical content and distilling key takeaways for busy decision-makers. I have experience creating clear and concise materials for elected officials, department heads, and community members, as well as presenting to city council committees and leading public workshops. I look forward to two-way conversations with community members to better understand their needs and incorporate community input into my policy recommendations.
I’m particularly interested in [City Name]’s work to [specific initiative, e.g., expand affordable housing, implement its climate action plan, make progress towards Vision Zero, etc.]. Whether by informing budget tradeoffs, conducting scans of best practices among peer cities, or evaluating the effectiveness of City programs, I’m interested in learning more about how I can contribute to this work. I’m interested in working in local government because I’m motivated by seeing immediate results from effective policy, whether it’s safer streets, increased access to stable housing, or more responsive services.
As a collaborative and deadline-driven professional, I am adept at managing multiple projects at once and enjoy working with a diverse range of people, including engineers, budget officers, legal staff, and nonprofit leaders. I am eager to bring my policy analysis, problem-solving, and communication skills to the City of [City Name], and I would welcome the opportunity to discuss my candidacy with you further. Feel free to contact me via [phone] or [email].
Sincerely, [Your Name]
Show the full output Show less
- GPTZero0.83passed
- Pangram1.00passed
- Winston AI0.97passed
- ZeroGPT0.60passed
- Copyleaks1.00passed
- Originality.ai0.06caught
Passed 5 of 6 detectors.
Every StealthGPT run in November 2026 →Each chip is one detector's score for that output, from 0 (AI) to 1 (human). A score of 0.50 or higher counts as a pass and shows green; below that the detector caught it and the chip shows red.
Penalties applied this cycle
How penalties work →The overall score shown above already reflects these deductions.
- Length deflation−1.0
The output came back much shorter than the input. The rewrite dropped content instead of paraphrasing it.
Compare with
Featured in
Watch the results
How the other humanizers did in the November 2026 test, each in under a minute.
Ryne AIwith Zoe Kopidis
Walter Writeswith Adrian Ahmetspahic
Undetectable.aiwith Ivan Jackson
WriteHumanwith Maxwell Bucholz
Humanize AI Prowith Ivan Jackson
SupWriterwith Zoe Kopidis
Humbotwith Adrian Ahmetspahic
Clever AI Humanizerwith Maxwell Bucholz
Rephrasywith Zoe Kopidis
Stealth Writerwith Adrian Ahmetspahic
Grammarlywith Ivan Jackson
HIX Bypasswith Maxwell Bucholz
Super Humanizerwith Zoe Kopidis
Phraslywith Maxwell Bucholz
AI Humanize iowith Ivan Jackson
Questions about StealthGPT
Answers are generated from the benchmark data and update every month. Last updated with the November 2026 cycle.
Is StealthGPT good?
StealthGPT ranked #2 of 16 in the November 2026 HumanizerBench cycle with an overall score of 69.47/100, above the field median of 61.14. On the parts of the score, it placed #5 of 16 on detector bypass, #12 of 16 on meaning preservation and #1 of 16 on readability. The score includes a 1-point deduction for length deflation.
Does StealthGPT bypass AI detectors?
In the November 2026 cycle, StealthGPT had a detector bypass rate of 66.0% (#5 of 16 tools). 3 of its 30 outputs passed all 6 detectors, and 27 were caught by at least one. Its average human-likelihood score from each detector, where 100 means every output was scored as human: Copyleaks 90.0, Winston AI 81.1, GPTZero 71.4, ZeroGPT 63.6, Pangram 60.9 and Originality.ai 28.8.
Can StealthGPT bypass GPTZero?
In the November 2026 cycle, StealthGPT's average human-likelihood score from GPTZero was 71.4 out of 100 (#6 of 16 tools against GPTZero). The highest average GPTZero score that cycle was 100.0, by Phrasly. See every tool's GPTZero results.
Is StealthGPT free?
No ongoing free humanizer plan is confirmed. The current introductory offer is $1 for 3 days, followed by the regular monthly Pro or Max rate. Paid plans start at $29.99 a month, billed monthly.
Is StealthGPT legit?
StealthGPT is a working AI humanizer at stealthgpt.ai. HumanizerBench has tested it in 6 monthly cycles, most recently November 2026, when it completed 30 of 30 test runs. Every input and output from its runs is published with each detector's verdict.
Does StealthGPT keep the original meaning?
StealthGPT scored 67.3 out of 100 for meaning preservation in the November 2026 cycle (#12 of 16; field median 72.8). The score measures how close each rewrite stays to the meaning of the original text. Its readability score was 77.7 (#1 of 16). It lost 1 point to penalties for length deflation.
Which kinds of writing does StealthGPT handle best?
In the November 2026 cycle, StealthGPT did best on blog posts (75.4/100) and worst on marketing copy (60.6/100). On academic essays it ranked #1 of 16.
How has StealthGPT ranked over time?
StealthGPT has been ranked in 6 HumanizerBench cycles: #7 in June 2026, #8 in July 2026, #12 in August 2026, #9 in September 2026, #2 in October 2026 and #2 in November 2026. Its overall score went from 71.47 in June 2026 to 69.47 in November 2026.
What are the best alternatives to StealthGPT?
In the November 2026 cycle, one tool ranked above StealthGPT: WriteHuman (#1, 70.11/100). Compare StealthGPT with any of them side by side.
Badges
StealthGPT is ranked in the November 2026 cycle, so it can embed any of these cards for free, and the live file re-renders every cycle, including when a rank falls. Usage rules →
Pinning a month for a press page? Use https://humanizerbench.com/badge/stealthgpt/november-2026.svg, which renders November 2026 forever.