People used to type a question into Google and pick from ten links. A growing number of them now ask AI instead and get one answer back. There is no page two in a chat window. If the model has never heard of your site, you were not beaten. You were never in the running.
Claude and GPT learned different things at different times. You always see both scores side by side.
Not one vanity number. We put eight separate questions to each model and show you every answer.
Models are not calculators. We ask three times and keep the middle answer, so noise never looks like a trend.
Search is switched off on every scan. We measure what a model already knows, not what it can look up.
A short list of what we do differently. Some of it is what we refuse to do.
Any domain, any time, with no account and no card. You only pay if you want a domain kept in your dashboard and re-scanned week after week.
No subscription and nothing to cancel. A domain slot is yours for good, which is why you will not find a billing page to fill out.
One model knowing you while the other has never heard of you is real information. A single blended number would have quietly hidden it.
A slot works on any domain, not only yours. Your score on its own is a number. Your score next to theirs is an argument.
Tracked domains are measured automatically, whether you are looking or not. The history is already waiting the day you want it.
We cannot promise to raise your score, because nobody outside the AI labs knows how training data gets picked. We say so before you pay, not after.
The short answers. If you want the reasoning behind them, the Why AIDRank page goes deeper.
A number from 0 to 100 showing how well Claude and GPT know your domain from memory alone. It comes from eight separate questions we put to each model, averaged into one figure. The score is the summary. The breakdown underneath it is where the useful part lives.
Because a model that can search will find any live website in a second and describe it back to you. That only proves your site exists, which you already knew. What matters is whether the model carries you around by default, since that is what decides if you get named when nobody mentions you.
Not for your score today. Nobody can rewrite what a model already learned in the past. But you can absolutely influence what the next model version learns about you: clear, well-structured content that gets cited by authoritative sources is exactly what shapes future training data. That is a real and growing practice, often called GEO. It just will not touch the number you are looking at right now.
Because a model's knowledge is frozen the day its training ended. Publishing something this week cannot change what a model already learned years ago. Scores move when a new model version arrives, every few months or so, and catching that jump is exactly what the weekly history is for.
Yes. Any domain, no account, no card. There are fair use limits per hour and per day so nobody can run thousands of scans in a loop, but normal use will never come close to them.
A domain slot. It puts one domain in your dashboard and keeps it re-scanned automatically every week, building a history you can look back on. You pay once, it never expires, and the slots stack if you buy more.
Yes, and most people do. A slot works on any domain, so plenty of them end up pointed at competitors rather than at your own site. Watching someone else climb while you sit still is usually the more useful chart.
No, and it does not replace it. Your Google ranking and your AI visibility have different causes and different fixes. Keep doing your SEO. This is the blind spot sitting right next to it, and almost nobody is looking at it yet.
One scan takes about thirty seconds. No account, no card, no commitment. You will know more about your AI visibility than most of your competitors know about theirs.
Scan My DomainHave a question about the score, the method, or your account? Write to us and we will get back to you.