Find out what ChatGPT, Gemini and Perplexity say about your company.
Eleven fixed prompts, five named engines, every answer quoted verbatim, including the parts that are wrong, missing or unflattering. The prompts are printed in the report so you can run them again yourself, any time, without us.
The AI Visibility Check is a free diagnostic, not software. A person runs the published prompt set against ChatGPT, Gemini, Perplexity, Claude and Google AI, records the model version and date, compares every answer to your real facts, and tells you which gaps a Citation Release can address, and which it cannot.
What you get
- What each of five engines states about your company today, quoted verbatim
- Every factual error and omission, itemised against your real details
- The eleven prompts, printed in full
- The engines and model versions queried, with the run date
- A plain statement of which gaps a release can address, and which it cannot
What it costs, and what it costs you afterwards
Nothing, and nothing. No call requirement, no card, no drip sequence. If you read the report and never speak to us again, the report was still the whole thing we promised.
The likely outcomes, stated in advance
One: the engines know you and state your facts correctly. You do not need us, and the report will say that. Two: the engines know you and get things wrong: the case a Citation Release is good at. Three: the engines have never heard of you: addressable, more slowly.
There is a fourth case we cannot sell you a fix for: the engines know you correctly but recommend someone else. Zero press-release domains appeared across 100 “best X” queries in Loganix’s March 2026 test.
Request your AI Visibility Check
Four fields. We reply to every enquiry, usually within one working day.
We store your enquiry to answer it. We do not sell, share or rent it, and we do not add you to a mailing list you did not ask for. Privacy.
Fair questions about a free AI visibility tool.
Is this an automated AI visibility tool?
No. The prompt runs are executed by a person against the live engine interfaces and captured verbatim. A person writes the summary and decides what a release can and cannot address. If that ever changes in a way that affects reproducibility, the method version changes and says so.
What is the catch?
The report is genuinely uncomfortable to read for most companies, and some people who read it buy something. That is the whole business model. It only works if the report is accurate, so making it accurate is in our interest.
Will you tell me a release won’t help?
Yes, when that is the answer. It is one of the four documented outcomes above, and step 1 of the process exists to catch it before you spend money.
Do you need access to anything?
No analytics, no Search Console, no admin credentials. Everything measured is a public AI engine’s response to a public prompt about a public company.
Can I just run the prompts myself?
Yes, and you should: they are printed in full in every report. What we add is the comparison against your real entity facts and a judgement about which gaps digital PR can close. If you would rather do all of that yourself, the printed prompts are enough.