Prices
- Source. The vendor's own pricing page. Never other review sites or aggregators: in October 2026 many ranking pages still quoted plans that vendors had already removed.
- Evidence. A dated screenshot and a text copy of each pricing page.
- Freshness. A script re-reads every pricing page each morning from a US server and flags changes. A person confirms each change on the live page before the site is updated, and it goes in the change log.
- Currency. We publish US dollar prices we have seen ourselves. If a page shows us another currency, we show that currency until a US check confirms the dollar figure.
- Cost per prompt. Monthly price divided by included prompts, as each vendor counts them. Cadence (daily or weekly runs) and engines covered differ between vendors, so we always show them alongside.
Hands-on tests (coming next)
AI assistants rarely give the same answer twice. A tracker that runs a prompt once can report noise as a trend. Our test is built around that:
- The same prompts in every tool, loaded on the same day, for a small set of test brands from large to tiny.
- Ground truth from many runs. Each prompt is run repeatedly in the assistants themselves, so we know how often a brand really gets mentioned, with a margin of error.
- Scores per tool: how close its reported visibility is to that ground truth, whether it names the same competitors and cites the same sources, whether it says if it reads the consumer app or an API, how fresh its data is, and what it costs per prompt.
We'll publish results, screenshots and the overall method. We keep the exact prompt list private so vendors can't tune for it.
Free trials only, for now: we test what anyone can try, or what a vendor gives us an account for, and the page will say which.
Independence
No vendor pays for placement, rankings or reviews, and there are no affiliate links on this site today. If that changes, every affected page will say so at the top, and rankings will still come only from data. See the about page.