StealthWriter AI detector

The free tier has two layers. The no-account demo: "Paste up to 500 words and select Scan for AI to see your results on this page. Daily demo limits apply", with no figure for the daily limit (stealthwriter.ai/ai-detector, read 10 November 2026). The Free plan at $0: "AI detector (V1 & V2)", "AI deep scan", "10 humanizations/day", "10 AI scans/day", "1,000 words per input" and the Ghost 5.2 Mini humanizer model (stealthwriter.ai/pricing, read 10 November 2026). The card records 10 checks a day in the free checks per day ranking, the no-account demo on the no-account view, and no monthly word figure, so StealthWriter sits in the second table of the free tiers ranking.

The pricing page states "All prices are in US dollars (USD)" and shows a Monthly and Yearly switch labelled "Save 17%"; the monthly prices are in the page text and the yearly prices come from the plan data embedded in the page's script, which divides each yearly price by 12 for the display (stealthwriter.ai/pricing, read 10 November 2026).

Every paid tier pairs its scan count with the same number of humanizations a day, and the page's one-line summary of the plans is about the humanizer models: "Free includes Ghost 5.2 Mini. Paid plans unlock Ghost 5.2 Pro and Legacy" (stealthwriter.ai/pricing, read 10 November 2026). The 5,000 words per input on paid plans is the figure StealthWriter carries into the maximum input per check ranking, a word figure stated by the vendor rather than one converted from characters.

Retention is the field where StealthWriter is clearest. The Privacy Policy, last updated 5 September 2026, lists among the data it collects "Text you submit for AI processing (humanization or detection). We do not store this text", and its Data Retention section states: "Content submitted for AI processing (humanization or detection) is not stored at all. It is classified by a moderation tool and then either rejected or processed. We do not use submitted text to train models" (stealthwriter.ai/privacy, read 10 November 2026). The card records zero days, the reading the index gave Sightengine's "deleted once it has been processed" and GPTKit's "immediately removed", and StealthWriter joins the cards at zero in the data retention ranking. Billing runs through Paddle or Stripe, and the policy says the company does "not store your full payment information".

StealthWriter publishes no accuracy figure and no false positive rate for its detector, so it has no row in the false positive ranking and no claim for the claim versus evidence ranking to pair. What the page states instead is a limit: "An AI detection score is an estimate, not proof of authorship", "It does not verify the author or establish which writing tools they used", and "Avoid using a detector score as the sole basis for academic or employment decisions" (stealthwriter.ai/ai-detector, read 10 November 2026). The FAQ repeats it: "Human writing can be flagged as AI, and AI-generated writing can go undetected. A score is an estimate, not proof of authorship."

On rewritten text the product's purpose runs the other way, and the vendor says so. The humanizer FAQ asks "Does humanizing guarantee a particular AI detector score?" and answers "No. Different detectors can give different results, and no rewrite guarantees a particular score" (stealthwriter.ai/humanizer, read 10 November 2026), and the terms state that StealthWriter "does not guarantee that content generated by the tool will be scored as human-written by AI detectors or plagiarism-checking software" (stealthwriter.ai/terms, read 10 November 2026). The disclaimer page adds that results "may vary based on the input text, selected settings, language, detector methodology, and updates made by third-party AI detection tools" (stealthwriter.ai/disclaimer, read 10 November 2026). The card records no statement that the detector catches paraphrased or humanized text.