ZeroGPT humanizer: what ZeroGPT publishes and the tools listed on one set of fields

The table lists eight tools in order of the largest request the cheapest paid plan or free tier accepts, largest first, because ZeroGPT scans up to 15,000 characters at a time and a tool that rewrites a document whole gives you one output to check rather than several joined. No column is a ZeroGPT reading. The ZeroGPT column of the published run will be added to this table when the first run publishes. Every figure comes from the vendor's own pricing or product page, checked on 8 November 2026.

None of those lines mentions ZeroGPT's reading, because none can yet.

ZeroGPT's DeepAnalyse page, read 9 November 2026, describes a multi-stage method that "moves from broad text-level analysis toward sentence-level and finer patterns before producing the final assessment", trained on internet text collections, educational datasets and ZeroGPT's own synthetic AI datasets produced with various language models (ZeroGPT, DeepAnalyse technology, read 9 November 2026). ZeroGPT's FAQ names the signals as token patterns, burstiness, entropy and ensemble classifier features, puts its accuracy at "pushing toward >98% on internal evaluations", and recommends analysing at least 150 to 200 words (ZeroGPT, FAQ, read 9 November 2026). The results screen highlights every sentence the model judges AI-written and shows a gauge with the percentage of AI inside the text, and the free checker accepts 15,000 characters per scan.

Three things about that table matter for a humanizer search. First, all three headline figures are ZeroGPT's own, measured on internal evaluations, and the page links no dataset, no date and no outside study behind them. Second, ZeroGPT publishes no rate for paraphrased or humanized AI text at all. Third, the gauge is a percentage of text judged AI, not a probability that the document is AI-written, and the two are different claims; what any detector's number means is set out in do AI detectors work?.

What is ZeroGPT's own humanizer, and is it independent?

ZeroGPT sells a humanizer beside its detector. Its AI Humanizer page, read 9 November 2026, offers a free tier, tells users to click "Check for AI" to see whether the rewrite is flagged by ZeroGPT's own detector, and offers an Advanced option that the page says is "provided by Undetectable AI". The page states the humanizer keeps the original meaning; it publishes no measurement of that.

A vendor's own detector is not independent evidence about that vendor's humanizer. When the same company sells the rewrite and the check, a human label is the company grading its own work, and the figure to look for is a reading from a detector the vendor does not own, on a dated run, with the input scanned before and after.

Why does this page rank nothing by a ZeroGPT reading yet?

Because a reading is only meaningful with a date, a version and a baseline, and this site has not published a run. The published method scans every output on five detectors, ZeroGPT among them, records ZeroGPT's own gauge reading on the run date with a screenshot, scans each input before humanizing so the change is visible, and counts "pass" only where the detector's own label is human. No detector score should be the only evidence in a misconduct case.

The peer-reviewed record explains the caution. Weber-Wulff and colleagues tested 12 public detectors plus Turnitin and PlagiarismCheck on human, machine, machine-paraphrased and machine-translated text and found that all 14 scored below 80 percent on their accuracy measure and that obfuscation such as paraphrasing lowered performance further (Weber-Wulff et al. 2023, International Journal for Educational Integrity, read 9 November 2026). ZeroGPT's current model was not in that study, and a 2023 study cannot speak for a 2026 reading, which is why a run carries a date.