For the complete documentation index, see llms.txt. Every page on this site is also served as Markdown: append `.md` to any URL, or send `Accept: text/markdown`.
Skip to content
Attensira LogoAttensira
Start free

Free tool

Answer citation checker

The answer citation checker puts one buying question to ChatGPT, Claude, Perplexity and Gemini with web search on, and shows every answer in full: the brands each one named, in order, and the pages it was built from. A run takes minutes, so it asks where to send the result. The report is a public link.

Start freeSee what AI says about you

7 days free · card required

Or use the free tool below

Ask your question

Minutes: four assistants' APIs with web search on · Free; the link is sent by email

Write it the way somebody choosing a vendor would — "best crm for a small sales team", not "is Attensira good". A question with a brand name in it measures nothing: the assistants will talk about the brand you named.

Four assistants, one question, every answer kept.

You get each answer in full, the brands each one named in the order it named them, and every page the answers were built from — grouped by site, so you can see which third-party pages decided the result.

It takes a few minutes, so we ask for an address and send you the link. The report itself is a public page you can share with anyone; nothing in it is behind the form.

Your result appears here

ChatGPT
who it named, in order
Claude
who it named, in order
Perplexity
who it named, in order
Gemini
who it named, in order
Pages cited
grouped by site

See our own run

Share
Perplexity app · Attensira trackingasked

Question: best AI search visibility tools

Perplexity answered: Here are top AI search visibility tools you can consider: SE Visible, Ahrefs Brand Radar, Profound AI, ZipTie, Similarweb, Petra Labs, Semrush AI Visibility Toolkit, OtterlyAI, Rankscale …

Sources cited

  1. 1visible.seranking.com
  2. 2zapier.com
  3. 3, 6semrush.com
  4. 4trustmary.com
  5. 5seoptimer.com
  6. 7seo.com
  7. +3 more

Not namedAttensira is missing from this Perplexity answer.

Stored answer, Perplexity sample 3 of 3, Attensira tracking · · n=39 answers, 27 Sep to 3 Oct

What does one run give you?

  1. The answers, verbatim

    Every reply from ChatGPT, Claude, Perplexity and Gemini in full, read through their APIs with web search on, so any claim can be checked against its text.

  2. Who was named, in order

    Brand names read out of each answer's own prose, in the order they appear. That order is the assistant's ranking.

  3. What the answers were built from

    Every page each assistant cited, grouped by site, with how many assistants leaned on each one.

What does an answer look like when we check our own name?

We track “best AI search visibility tools” for our own brand: a buying question with no brand name in it, the kind this checker is for. These rows come from that paid tracking of the ChatGPT and Perplexity consumer apps, 27 September to 3 October 2026. Tracking samples up to three answers a day per engine and leaves out failed runs, so the week holds 21 ChatGPT and 18 Perplexity answers (n=39). The free checker reads the model APIs once each and reports the same three things: who was named, in what order, and which pages the answer was built from.

ChatGPT · Attensira named
0 of 21 answers— Fails
Perplexity · Attensira named
0 of 18 answers— Fails
Named most often
Semrush in 30 of 39, Ahrefs in 29 of 39
ChatGPT · built from
aiattention.ai and ayzeo.com, each in 21 of 21
Perplexity · built from
trustmary.com in 12 of 18, semrush.com in 11 of 18
attensira.com cited
0 of 39 answers— Fails

Seven days, 39 answers, two engines: Attensira is named in none of them, Semrush and Ahrefs in most, and no page of ours is cited. That is what the checker exists to surface. For a question with no brand in it a new product starts at zero, and the fix is on the third-party pages the engines read: off-site work, not a page edit.

The answer at the top of this page is the stored answer, Perplexity sample 3 of 3, Attensira tracking, asked on 1 Oct 2026. The counts above cover 27 Sep 2026 to 3 Oct 2026.

Is this what a person sees in ChatGPT?

We read model APIs with web search on. That is not the same surface as the consumer apps a person types into (different system prompts, different retrieval, sometimes a different model version), and a brand can swing sharply between the two on the same day, on the same question.

So the report labels the surface next to the numbers rather than in a footnote, and we do not average the two. If what you need is what the consumer products say, that is what the paid product reads; this is a real measurement of a real surface, and it is not that one.

Why one answer per assistant?

Each assistant is asked once. Running the same prompt repeatedly returns different citations, so a rate computed from a handful of answers is a snapshot rather than a settled fact, and a free tool cannot afford the depth that would make it one.

The honest response is to carry the number of answers everywhere the percentage goes, and to refuse to rank inside the noise. At four answers, a one-place difference in average position is not a finding, and the report says that on the page rather than leaving you to infer it. Sampling once and writing it up as though it were settled would be the same overconfident reporting this product exists to argue against.

What does the crew do when an engine reads pages you are not on?

Inside Attensira your buyer questions are asked on a schedule, at most once a day per prompt, engine and country, and every answer is kept verbatim. When an engine leans on pages you are not on, the write-up names them, and the off-site work is drafted for you to approve. What Attensira reads, and how often.

Inbox1 waiting for you
  • Researcher agentApproveYour category answers are built from pages you are not on: the three to approach first

    From your tracked answers · with the sample size

    Names the roundups and review sites ChatGPT and Perplexity leaned on for the question, and how often each named you. Drafted: what to pitch to each, for you to approve before anything is sent. Your choice: approve, edit then approve or decline.

Researcher agentResearcher · would hand you an inbox row like thisExample · the kind of change this finding calls for, not one that shipped

Why does it ask for an email address?

This one costs us real money per run and takes minutes: several live web searches and an analysis pass. So we ask for a name and an address before starting one, and send you the link when it is done rather than making you sit on a loading screen.

Nothing in the result is behind that form. The report is a public page with its own permanent link, shareable with anyone, and it stays that way. The exchange is for the running of it, not the reading. Every other tool in the directory answers immediately and asks for nothing at all.

The newsletter tick beside the address is separate and starts unticked. Asking for a report is not agreeing to hear from us again, and the two are separate lists you can leave separately.

What do you do with what it shows?

Being named and being cited are different facts with different fixes, so the report keeps them apart. Each pattern points somewhere else.

What it showsWhyWhat to do
Your page is cited, and the answer names a rivalWhy: The page was good enough to retrieve and not good enough to convince.What to do: A content fix on that page: answer the question it was cited for, directly and early. Find the passages an assistant can lift
You are named, but none of your pages is citedWhy: The assistant knows you from other sites. What it says about you comes from their pages.What to do: Read the cited pages: they are where your description comes from, and where it is corrected.
You are not named, and the cited pages are not yoursWhy: The pages the answer was built from do not mention you, or say too little to compare.What to do: Off-site work: get onto the pages it cites. Nothing on your own site changes this answer.
The assistants disagreeWhy: Each searched different pages. One answer each is a snapshot, not a rate.What to do: Do not rank inside the noise. Ask again on another day before acting on a one-place gap.

Questions people ask about this check

Is this what ChatGPT told a real user?

No. It is what the model behind ChatGPT answered our request through its API, with web search on and a system prompt that mimics the product's answering style. That is a real answer from a real surface, and it is not a recording of somebody's session. Nobody outside OpenAI can show you that.

Why did my brand not appear at all?

Sometimes because it was not named, and sometimes because the question was not one your buyers would ask. A question with no commercial intent, such as "what is a CRM", returns an explainer that names no vendors and measures nothing. Try the shape a buyer types: "best X for Y".

Can I run the same question again?

Yes, and after a week it re-runs and updates the same page at the same link. Inside that window you get the report we already have, immediately. That is deliberate: the answer to your question already exists, and re-running it would spend money to produce roughly the same page.

Who can see my report?

Anyone with the link, which is unguessable and only goes to you until you share it. The page shows the question and the answers; it never shows who asked for it. Your address is used to send you the link and is not on the page.

How are the findings written?

From the same rows you can see, under published instructions: the report links the exact version of each one it used. A finding that does not point at evidence on the page is dropped before the report is stored, so there is no commentary here that you cannot check.

This asks one question once. Attensira's crew works on the answers.

Start freeSee what AI says about you

7 days free · card required