About This Community Hub

Our Mission

WalterWritesAI.com exists to give students, educators and writers a clear, sourced picture of Walter Writes AI and the detectors it is measured against. The official product site explains what the tool does. This guide explains how to use it well, what it costs, where it fits and where the rules draw the line.

Three commitments shape every page. We answer the question in the first paragraph. We back every factual claim with a source you can open. And we tell readers what their course, employer or client rules mean before they paste a single word into any tool.

Who This Guide Is For

Most readers arrive with one specific question. This table points each group to the page that answers it fastest.

How We Research and Update Guides

Every guide follows the same process, from first draft to each later revision.

1. Start from primary sources. Vendor help centers, university policy pages, style guides and peer-reviewed or preprint research come first. The SPJ Code of Ethics puts it simply: "Use original sources whenever possible."
2. Check product facts on the vendor's own pages. Walter's plans and features come from walterwrites.ai and its help center; detector limits come from the detector's documentation or a university guide that reproduces it.
3. Date the evidence. Pages show when they were last updated, and comparison tables state when vendor pages were read.
4. Re-check on a schedule. Pricing, free tiers and detector rules change often, so the pages that depend on them are reviewed first in each update cycle.
5. Record what changed. Substantive revisions get a dated scope or revision note on the page itself.

Detector documentation shows why dates matter. According to the University of Melbourne's Turnitin guide, Turnitin changed its AI Writing Report on July 8, 2024 so that scores below 20% show only an asterisk. A guide written before that date would describe the report incorrectly today.

Sourcing Standards

• Inline links to the source. A statistic, policy, price or date links to the page it came from, next to the claim.
• Short, exact quotations. Quotes are copied word for word from the source and kept brief.
• No invented numbers. If a figure cannot be traced to a source, it does not appear on the site.
• Vendor claims are labelled as vendor claims. What a company says about its own product is kept separate from independent research.
• AI-suggested references are checked by hand. The APA Style team reported that one of five references ChatGPT supplied did not appear to be a real article, so no reference reaches a page without being opened.

How We Judge a Rewrite

A humanizer is judged on the text it produces, not on its marketing. Every evaluation on this site asks the same five questions, in this order:

1. Meaning. Does each sentence still make the same claim, with the same certainty?
2. Facts. Are names, numbers, dates and quotations unchanged?
3. Sources. Does every citation still support the sentence beside it?
4. Readability. Is the rhythm varied and the tone right for the reader?
5. Detector context. Which detector, which date, and what does its own documentation say about its limits?

Detector results come last on purpose. A rewrite that scores well but changes a quotation is a failed rewrite. For worked examples, see the Walter rewrite examples.

What We Do Not Publish

• Promises that any tool makes text permanently invisible to every detector.
• Statistics, quotes or study results that cannot be traced to an open source.
• Reviews written to order for a vendor, or rankings that change with commission rates.
• Advice to submit AI-assisted work where a course or employer prohibits it.

Corrections Policy

Errors get fixed fast and in the open. The SPJ Code of Ethics asks publishers to "Acknowledge mistakes. Correct them promptly and prominently." That is the standard this site applies.

A correction fixes something that was wrong: a misstated price, an outdated limit, a misread study. The corrected page gets a new updated date, and substantive corrections are described in the page's revision note. An update adds new information, such as a changed free tier or a new detector policy, and is dated the same way.

When a vendor's own pages disagree with each other, the page shows both figures and says so, rather than picking the more flattering one.

How We Cover Academic Integrity

Course rules come before any tool. The University of Toronto, for example, states that students may not use generative AI in a course unless the instructor explicitly permits it. Guides on this site tell students to check the syllabus and assignment instructions first, and to disclose AI use wherever it is allowed and required.

Detector scores are reported for what they are: probability estimates. Turnitin's own guidance says its AI report should not be the sole basis for adverse action against a student, and Liang et al. (2023) found widely used detectors consistently misclassify non-native English writing as AI-generated. Both facts appear wherever a page discusses detector results.

Frequently asked questions