Does Ryne AI bypass Pangram? Nobody has tested it. Ryne publishes a bypass percentage for ten named detectors and has reviewed six more on its own blog — and Pangram appears in none of them.
The only tool that bypassed Pangram and Turnitin in 2026.
Most humanizers clear one detector and get caught by the other. StealthWriter is the one that gets past both — run Ghost 5.2 Pro at level 7–8, section by section, and re-check before you submit.
Try StealthWriter →Key Takeaways
- Ryne AI publishes a claimed bypass rate for ten named detectors — Turnitin 100%, GPTZero 99.8%, Originality.ai 99.8%, Copyleaks 99.7%. Pangram is not one of them.
- Its blog carries roughly 31 competitor and detector teardowns, including one claiming a 100,000-text study on GPTZero. Pangram appears in none of them.
- Pangram has never tested Ryne AI either — absent from the DAMAGE paper, the 20-tool benchmark, and the third-party evaluations page.
- The only Ryne-versus-Pangram evidence anywhere is one screenshot in an affiliate-disclosed Medium post, with no source text, length, model or repeat trial.
- No methodology, sample size or test date is published for any of Ryne’s four percentages.
- Refunds require all of: within 3 days, under 500 coins used, under 1 AI report — then a 30% processing fee. Renewals are non-refundable.
- The honest answer is nobody has tested it. That is not the same as saying it works.
The only tool that bypassed Pangram and Turnitin in 2026.
Most humanizers clear one detector and get caught by the other. StealthWriter is the one that gets past both — run Ghost 5.2 Pro at level 7–8, section by section, and re-check before you submit.
Try StealthWriter →Does Ryne AI bypass Pangram?
Unknown, and the gap is conspicuous rather than accidental. Most tools in this market are silent about detectors generally. Ryne is the opposite: it is one of the most quantified, most publication-heavy vendors in the category. It names ten detectors, attaches a percentage to four of them, and has published dated teardowns of six.
Pangram is in none of it.
This page reports no first-party test and reviews the published record only. Every claim is dated and attributed, and where a source has a commercial stake — including our own — we say so.
What does Ryne AI actually claim?
Its homepage leads with “Bypass Turnitin and GPTZero with a 99.9% success rate”. Its humanizer page repeats the headline figure and breaks it down per detector.
| Detector | Ryne’s claimed bypass rate | Methodology published? |
|---|---|---|
| Turnitin | 100% | No |
| GPTZero | 99.8% | No |
| Originality.ai | 99.8% | No |
| Copyleaks | 99.7% | No |
| Winston AI, ZeroGPT, Writer, QuillBot, Crossplag, Content at Scale | named, no figure | No |
| Pangram | not listed anywhere | — |
Ten detectors are named across those pages: Turnitin, GPTZero, Content at Scale, Copyleaks, Originality.ai, Winston AI, ZeroGPT, Writer, QuillBot and Crossplag. Pangram is not among them.
Nor is any methodology. There is no sample size, no test corpus, no date and no model named behind the 100%, the 99.8%s or the 99.7%. The supporting material is testimonial copy — “passed Turnitin with flying colors”, “My professor couldn’t tell the difference”. Our full Ryne AI review covers the product itself, and Ryne AI vs Undetectable AI covers the comparison.
Why does the omission matter?
Because Ryne is not a company that avoids naming detectors. Its blog runs roughly 25 competitor reviews — StealthGPT, Undetectable.ai, HIX Bypass, Humbot, StealthWriter, GPTHuman, Walter Writes and more — plus five dedicated detector reviews of Copyleaks, Sapling, ZeroGPT, Winston AI and Turnitin.
One of those posts is titled “Why GPTZero is not reliable anymore. We Ran 100,000+ Texts to prove it.” Another asks directly whether Ryne can beat Turnitin’s latest update. A third is headed “Does Ryne AI Bypass GPTZero?”
That is not proof of anything. It is a pattern worth naming, and it is the most defensible observation available on this question.
Has Pangram ever tested Ryne AI?
No. Ryne appears nowhere in the DAMAGE paper, which audited 19 humanizing tools by name; nowhere in the August 2025 benchmark of 20 tools; and nowhere on Pangram’s third-party evaluations page.
And since July 2026 the question has become unanswerable from outside. The Pangram 4 technical report tests 13 commercial humanizers and labels every one “Commercial A” through “Commercial M”. Ryne may be in that set. It may not. Nobody outside Pangram can say, and anybody who claims Ryne is a specific letter is guessing.
What about the one test that exists?
There is exactly one artefact showing Ryne output against Pangram: a screenshot inside a March 2026 Medium round-up, captioned as showing fully humanised content. It is worth being specific about why that is not evidence.
The article discloses affiliate links. The screenshot is a single sample — no source text published, no word count, no model named, no repeat trials, no control. And the same article rates Pangram 9.2 out of 10 for accuracy while displaying it being beaten, which is internally incoherent.
One screenshot from a page that earns commission on the tool it is testing does not settle a question this market has an enormous incentive to answer. We apply the same standard to ourselves — see the disclosure at the top of this page.
What do users actually report?
Very little, and the little there is points somewhere unexpected. Ryne is not a tool Reddit organically discusses as a humanizer — the recommendations we found came from a small number of accounts posting the same suggestion across unrelated subreddits, all scoring between zero and two.
The two most plausibly organic comments are both about Ryne’s AI checker rather than its humanizer: a student reporting they had run roughly 100,000 words through it across about twenty assignments, describing it as close to Turnitin within about five percentage points. That is a different product doing a different job.
Zero Reddit posts or comments link Ryne AI to Pangram in either direction.
What should you take from this?
One thing only, stated precisely: the question has not been answered. Not by Ryne, not by Pangram, not by any independent party. Anyone telling you Ryne beats Pangram — including anyone earning commission on it — is going beyond the record.
What the record does support is Pangram’s own published ceiling for the category: across 13 unnamed commercial humanizers it reports catching humanized text 97.69% of the time, with the best single performer still caught 92.78% of the time. Whether or not Ryne is in that set, that is the realistic shape of the problem.
If your real worry is being wrongly flagged on your own writing, keep your drafting history — it costs nothing and outperforms any rewriting tool as a defence, and it matters most for the writers detectors treat worst: ESL writers and AI detection and false positives for neurodivergent students. Check before you submit with the best pre-submission check and the Turnitin self-check, and strip the obvious tells by hand first via what to remove before using an AI humanizer.
For the Pangram cells where real evidence exists, see Undetectable AI, StealthGPT, GPTHuman, WriteHuman, QuillBot, StealthWriter, Phrasly, Humbot and Rephrasy, plus the Pangram review.
The one tool that got past both
Pangram and Turnitin catch most humanizers — one or the other, and usually the one you did not think to test. StealthWriter is the only tool that bypassed both in 2026. Run Ghost 5.2 Pro at level 7–8, section by section, and re-check before you submit.
Free plan to start, no card. Or get the free prompt pack first →
Affiliate link — I earn a commission if you upgrade.
Frequently asked questions
Does Ryne AI bypass Pangram?
Nobody has tested it. Ryne publishes no Pangram figure, Pangram has never tested Ryne, and the only thing resembling evidence is a single screenshot in an affiliate-disclosed Medium article. The honest answer is unknown — not yes.
What bypass rates does Ryne AI actually claim?
On its humanizer page: 100% against Turnitin, 99.8% against GPTZero, 99.8% against Originality.ai and 99.7% against Copyleaks, under a headline figure of “99.9% Bypass Success Rate.” No methodology, sample size or test date is published for any of them.
Why isn’t Pangram in that list?
Ryne has not said. Ten detectors are named on its humanizer landing page and Pangram is not among them, despite Pangram being the detector independent benchmarks rank hardest to beat.
Hasn’t Ryne reviewed other detectors?
Extensively. Its blog carries dated teardowns of Copyleaks, Sapling, ZeroGPT, Winston AI, GPTZero and Turnitin — including a post claiming it ran more than 100,000 texts to assess GPTZero. Across roughly 31 competitor and detector posts, Pangram is mentioned zero times.
Has Pangram ever tested Ryne AI?
No. Ryne does not appear in Pangram’s DAMAGE paper, which audited 19 tools, nor in its August 2025 benchmark of 20 tools, nor on its third-party evaluations page.
What about the screenshot showing Ryne passing Pangram?
It comes from a Medium article that discloses affiliate links. It is a single screenshot with no source text, no word count, no model named and no repeat trials — and the same article rates Pangram 9.2 out of 10 while showing it being beaten. Treat it as an anecdote.
What does Ryne AI cost, and can I get a refund?
A free tier gives 100 coins and 250 words per go, English only. Paid plans run $29.99, $39.99 and $109.99 monthly, cheaper annually. Refunds require a request within 3 days with under 500 coins and one AI report used, and carry a 30% processing fee. Renewals are non-refundable.
Who operates Ryne AI?
Ryne AI L.L.C, licence number 2528619, registered in the Meydan Free Zone, Dubai, UAE. Its terms are governed by Meydan Free Zone rules and UAE federal law.
