Editorial Policy

This page explains how we research, write, label and correct the content on this site.

How we research

• We inspect the live public product interface, official product pages, terms of service, privacy policies and content policies.
• We record the date each page was checked and show it as a “last checked” badge.
• Prices, token allowances and feature gates are quoted as reference figures with an explicit note that the live checkout is final.
• Third-party observations are attributed as reports, not presented as our own testing.
• We do not invent generation counts, success rates, testimonials or testing volume. Where a benchmark has not yet been run, we say so.

How ratings work

Editorial scores reflect the inspected public workflow, documented features, price-to-feature ratio and policy checks. They are opinions grounded in that method, not aggregated user reviews. We do not publish AggregateRating markup because we do not collect user-review data.

Affiliate independence

Affiliate relationships (see the affiliate disclosure) never determine verdicts, rankings or safety guidance. Comparison pages use fixed criteria stated on the page, use conditional verdicts and label affiliate landing pages as such.

Safety standards

Every guide follows our safety policy: consent guidance appears before operational steps on sensitive pages, all imagery stays non-explicit, and we never publish prompts or workflows involving minors, real non-consenting people, coercion or deception.

Updates and review cycle

• Commercial pages (pricing, coupon) are re-checked when the partner or platform announces changes, and periodically otherwise.
• Each article displays its most recent check date.
• Superseded claims are corrected in place; substantive corrections are noted on the page.

Corrections

If you find a factual error, an outdated price or a broken link, email [CONTACT_EMAIL] with the page URL and the issue. We review correction requests, verify against primary sources and update the page. Requests about unlawful or non-consensual content follow the faster content-removal route.

Use of AI tools

Drafting and production tooling may include AI assistance. Every published claim is constrained to the sourced material described above, and a human editor is responsible for the final page.