Editorial policy

How tools are chosen

• Editors add tools that people actively search for (measured with keyword data for English and Indonesian searches) and that we can verify from official sources.
• Makers can submit their own tool. A submission is reviewed by an editor before it appears, and a maker can only claim a listing after proving control of the tool's domain.
• Every fact on a listing (price, free plan, Indonesian support, features) links to the official page it came from and shows the date we checked it. Facts older than 90 days are marked "needs re-check" and left out of comparisons.
• A weekly checker visits every tool and its pricing page. A changed pricing page is marked "needs re-check"; a tool that fails three checks in a row is marked inactive and removed from lists, rankings and the sitemap.

Sponsor placements

• Any paid placement shows a visible Sponsor label next to the tool, and its link uses rel="sponsored".
• Sponsor placements sit in clearly marked slots. They never change the order of an organic list, a ranking or a benchmark table.
• We do not sell backlinks, "dofollow" links, reviews, votes or scores.
• Paid packages are not open yet; prices are shown as proposals on the advertise page.

Affiliate links

We have no affiliate links today. If that changes, every affiliate link will be marked on the page where it appears, following the FTC Endorsement Guides (16 CFR Part 255), and this page will say so before the first one goes live.

Products made by our owner

The owner of Newly Stack also builds AI products (listed on the about page). They carry an ownership label everywhere they appear, will be tested with the same public method (none of them is tested yet), and are never ranked higher because of who made them. In week-long launches they are listed apart from the ranking.

Clients of our owner's agency

The owner also runs Creativism, a digital marketing agency. If a tool made by a Creativism client is ever listed, it carries a "Client of our owner's agency" label on every card, profile and launch, and the same test and ranking rules apply.

Tests and scores

Scores come from tests we run ourselves, dated and versioned, with prompts and raw outputs stored. Writing tasks are graded by AI judges from vendors other than the tested model's vendor and are labeled "model-graded" until human raters review them. See how we test and the benchmark methodology.

Corrections and disputes

• Every change to a published score is recorded in the public score change log with a written reason. A database rule rejects score changes without a log entry.
• Verified makers can dispute a score within 30 days of its publication. An editor aims to answer within 7 days with a correction, a re-test or a reasoned decision to keep the score, and the maker may publish a response of up to 500 characters.
• Brand owners can ask for a listing change, a logo removal or a correction through the legal request form; we aim to answer within 5 business days of email confirmation.

Use of AI

We use AI models as test subjects and as benchmark judges, and some of our texts are drafted with AI help. Every published fact still needs an official source and a check date, and our own test results are labeled as ours.