StealthWriter AI Humanizer Review: Limits and Pricing
StealthWriter combines a browser humanizer with a no-account detector. The useful questions are how its daily actions, input cap, free follow-ups, billing, and text workflow fit the job you need to finish.
HumanPen Team
· 28 min read
Is StealthWriter worth trying?
StealthWriter is a practical option when your input is plain text, it fits the plan's per-request limit, and you want several rewrite strengths plus a built-in detector. Its public pages do not establish an editable document-file round trip, a public API, a browser extension, or a future Turnitin result.
I opened the signed-out homepage, FAQ, payment and refund policies, privacy and terms pages, changelog, sitemap, and likely product routes. The date was August 26, 2026. Every input stayed blank; I remained signed out, uploaded nothing, and never opened checkout.
The publisher is HumanPen. Its document workflow appears near the end. Both vendors speak for their own products here; neither set of claims is treated as neutral benchmark evidence.
StealthWriter's most useful public details are not pass-rate claims. They are the two usage meters, the follow-up actions, and the shape of the result.
How do StealthWriter's limits and prices work?
Every plan has at least two limits to track: how many initial actions it allows each day and how many words one input can hold. The homepage publishes monthly prices and daily scan counts through Pro. The FAQ adds annual prices and the Scale plan.
| Plan | Public price | Initial humanizations per day | AI scans per day | Words per input |
|---|---|---|---|---|
| Free | $0 | 10 | 10 | 1,000 |
| Starter | $20/month or $200/year | 50 | 50 | 5,000 |
| Plus | $50/month or $500/year | 150 | 150 | 5,000 |
| Pro | $100/month or $1,000/year | 350 | 350 | 5,000 |
| Scale | $400/month or $4,000/year | 2,000 | Not stated on the checked pages | 5,000 |
An April 25, 2026 Changelog entry also lists a Business plan with 800 humanizations per day. Business does not appear in the current homepage or FAQ plan list, and the Changelog does not give its price or scan allowance. The public pages checked here do not show whether Business was discontinued, remains available inside an account, or was omitted from the current list. The table therefore reports the current homepage and FAQ rather than claiming to be a complete catalog.
The FAQ says daily limits reset at midnight UTC. It also says Free usage is subject to fair-use limits that may vary, even though the homepage and plan answer publish ten daily humanizations. If ten runs matter to your schedule, verify the number shown inside the account that day; the card alone is not enough.
A larger daily allowance does not increase the paid-plan input above 5,000 words per request.
What happens after the first Humanize action?
StealthWriter separates the first Humanize action from several ways to revise its result. According to the FAQ, only the initial action reduces the daily humanization allowance.
| Action | Starts from | Effect on the published daily humanization count |
|---|---|---|
| Humanize | Original input | Uses one initial action |
| Rehumanize | Original input again | Does not reduce the count |
| Humanize More | Current output | Does not reduce the count |
| Re-humanize Output | Current output again | Does not reduce the count |
| Rescan result | Current output | Does not reduce the humanization count |
The humanizer offers levels 1 through 10. The FAQ groups 1 to 3 as light changes, 4 to 6 as balanced, and 7 to 10 as heavier rewrites. Free uses Ghost 5.2 Mini, while paid access includes Ghost 5.2 Pro. These are vendor labels, not quality results reproduced here.
After a result appears, the FAQ says a user can click a sentence and choose from up to three alternatives. The page gives no numeric follow-up allowance, and none is added here.
Count the first run, the follow-up controls, and the input size separately. They answer different capacity questions.
What does the built-in detector measure?
The signed-out homepage exposes a separate detector with a 500-word meter and no account requirement. Users can choose V2, the default, or V1. Deep Scan is described as sentence-level color highlighting.
| Detector detail | What the public page states | What this review verified |
|---|---|---|
| Input size | 500 words on the homepage | The empty meter displayed `0/500` |
| Score direction | Percentage classified as natural | No text was scanned |
| Models | V2 default, V1 available | Both controls were visible |
| Deep Scan | Sentence-level highlighting | The feature description was visible |
That score is StealthWriter's own naturalness estimate. It is not a reproduced Turnitin AI Writing score. The FAQ calls V2 more accurate than V1, but no validation set, detector agreement test, or named Turnitin result was reproduced for this article.
Use the detector as one product control, not as proof of how another detector will classify a submission.
What goes in and what comes back?
The public workflow checked for this review is text-first. The homepage humanizer link goes to sign-up, while the open detector uses a textarea. The FAQ presents pasted text as the starting point, then documents text alternatives in the result.
| Public evidence | What it supports | What it does not establish |
|---|---|---|
| Homepage textarea and no file input | A paste-text entry point | Successful DOCX, PPTX, or PDF upload |
| Markdown warning and Remove Markdown option | Formatting may be stripped before a run | Preservation of Word styles, fields, or page layout |
| Up to three sentence alternatives | Editable choices inside the web result | A same-format editable file download |
| No download control before sign-in | No public export route was visible | Every possible account-only output state |
The FAQ recommends plain text for best results when headings, bold text, links, or code blocks are present. That advice is useful for a text workflow, but it also means the public evidence does not support a formatting-preservation promise.
That conclusion stops at the surfaces I could inspect. An account screen could expose something else; the signed-out homepage and FAQ did not show an editable file returning in either DOCX or PPTX format.
A result you can edit in a browser is not automatically the editable document you must submit.
What is available outside the web page?
StealthWriter's FAQ is unusually direct about its current access surface. It describes a browser product rather than a set of integrations.
| Surface | Current public statement |
|---|---|
| Web | Works in modern desktop, tablet, and mobile browsers |
| API | No public API at this time; one may be offered later |
| Browser extension | Not available yet; listed as a future roadmap item |
| Languages | Spanish, French, German, Italian, Portuguese, Dutch, and others are named |
The footer identifies the service as StealthWriter by AiVantage LLC. This review does not transfer features from StealthGPT or any similarly named product.
If an integration is essential, verify the current access route before buying. A roadmap item is not a usable endpoint.
What billing conditions matter before payment?
Price is not the whole purchase condition. StealthWriter's public FAQ and policies also spell out renewal, refunds, and plan changes.
| Condition | Public wording summarized |
|---|---|
| Free entry | $0 and no credit card required |
| Paid trial | No duration, introductory price, or conversion schedule was published on the checked sales pages |
| Renewal | Paid subscriptions renew automatically unless cancelled before renewal |
| Refund | Standard window: request within 24 hours of purchase or renewal if the account was not used during that period. Exceptional circumstances may be submitted for case-by-case review, but approval is solely at AiVantage LLC's discretion |
| Upgrade | Prorated and charged immediately |
| Lower paid plan | Takes effect immediately, with no refund for unused time |
| Move to Free | Cancellation is scheduled for the end of the current billing period |
The standalone `/pricing` route returned 404 during this check. The current numbers came from the homepage and FAQ. The FAQ uses the account-state word `trialing` in a plan-change answer, but no paid-trial offer or terms appeared on those pages.
Reopen the FAQ and policies before payment. Annual prices and renewal rules can change independently of the homepage cards.
Which jobs fit the public contract?
StealthWriter's checked pages support some jobs more directly than others.
| Job | Fit based on public evidence | Reason |
|---|---|---|
| Rewrite plain text under the plan limit | Direct fit | The FAQ documents paste input, levels, and sentence alternatives |
| Check up to 500 words without an account | Direct fit | The homepage exposes the separate detector |
| Process a 12,000-word paper in one run | Not supported by the published limit | Paid inputs stop at 5,000 words |
| Return the same editable DOCX | Weak public evidence | No checked page documents that round trip |
| Call the humanizer from an agent or application | Not currently documented | The FAQ says there is no public API at this time |
| Guarantee a future Turnitin result | Not supported | No Turnitin benchmark was run or documented here |
Before using a real draft, test a disposable passage. Include a number, a citation, a cautious qualifier, and a specialized term whose wording is fixed. Save the source, try the rewrite level you expect to use, and compare every claim before accepting the result.
The Terms and FAQ put compliance with workplace or institutional AI rules on the user, including avoiding misrepresented authorship. Rewriting does not remove disclosure duties from the final work.
Test the exact input, rewrite level, and output handoff you plan to use, not only the product's sample text.
When is a document workflow more relevant?
When the required output is an editable office file, such as DOCX or PPTX, ask for evidence at both ends: which file enters, which file returns, and which passages are allowed to change.
The HumanPen document workflow accepts DOCX and PPTX documents, plus TXT input. The product page states that the returned file keeps its input type and lets the user define scope by passage or from an imported Turnitin report. This article is produced by HumanPen, so these are vendor claims rather than a shared test. Check the returned file yourself.
Eligible results can continue lowering AI for free.
Frequently asked questions
Is StealthWriter free? It has a Free plan listed at $0 with no credit card required. The homepage publishes ten initial humanizations and ten scans per day with 1,000 words per input, while the FAQ separately warns that Free usage is subject to fair-use limits that may vary.
What is StealthWriter's word limit? Free is listed at 1,000 words per humanizer input. Paid plans are listed at 5,000 words per input. The no-account homepage detector has its own 500-word input.
Do the free follow-up actions use another daily humanization? The FAQ says Rehumanize, Humanize More, and Re-humanize Output do not reduce the daily humanization count. It does not state a numeric follow-up allowance.
Does StealthWriter upload and return DOCX files? The signed-out homepage had no file input, and the FAQ documents a paste-text workflow with optional Markdown removal. The checked public pages did not establish an editable DOCX round trip. Account-only states were not tested.
Does StealthWriter have an API or browser extension? The current FAQ says there is no public API and no browser extension yet. It describes both as possible future additions, not current access routes.
Will StealthWriter pass Turnitin? This review cannot support that conclusion. No text was humanized, no detector was run, and StealthWriter's own naturalness score is not a Turnitin report.
Official pages checked
The list below records the public sources used on August 26, 2026. Check them again before relying on a plan or workflow because controls, prices, and policies can change.
KEEP READING