Home · AI visibility audit
AI visibility audit: 11 buyer questions, 5 models, one report
An AI visibility audit asks a fixed set of buyer questions once, scores who was named, and hands you a report you can send today.
That is a different purchase from a platform that watches a prompt list every day. If you need a run every day or every week without opening this page, this audit is the wrong tool. If you need to show a teammate where ChatGPT, Claude, Gemini, Perplexity, or DeepSeek left you off a shortlist, it is the right one.
What this check is
- One run you start. There is no scheduled run.
- No SeenByLLMs account, no shared seats, no data that follows you to another machine.
- You bring an OpenRouter key. Model calls are billed by OpenRouter, on top of $0, $12/mo, or $49 once.
- API answers, not the ChatGPT website and not Google AI Overviews.
- It reports what the models said this time. It does not promise a higher rank.
What the published audits contain
Each case study used 11 questions written the way a buyer types them, not the way a marketer phrases a keyword. A brand counts as named when the answer contains the brand name or an alias as plain text. Rank is the order brands first appear. Pitch targets are cited pages that name a rival and not you.
- Plausible — absent from 4 of 11. Score 61. Share of voice 33% vs Matomo 19%.
- Buttondown — absent from 3 of 11. Score 52. Share of voice 18% vs ConvertKit 23%.
- Tally — absent from 3 of 11. Score 57. Share of voice 21% vs Typeform 24%.
- Cal.com — absent from 2 of 11. Score 70. Share of voice 39% vs Calendly 35%.
- Raycast — absent from 2 of 11. Score 67. Share of voice 28% vs Alfred 25%.
Those five runs used one free model with web search, about $0.08 each. They are not customer results, and they are not a promise that your audit will look like theirs. Your audit uses the rivals and questions you enter.
Score, in one line
Score = round(100 × (0.5 × mention rate + 0.3 × share of voice + 0.2 × rank score)), where rank score is 100 for first place down to 0 at sixth or worse. Failed calls are left out. The formula is published because a secret index is not something you can forward to a teammate.
Questions
How is an AI visibility audit different from monitoring?
An audit is one pass over the questions you care about. SeenByLLMs does not run it again tomorrow unless you open the page and press Run.
Why 11 questions?
The five published runs on this site each used 11 buyer questions, one free model with web search, on 16 October 2026. In the app you can use fewer or more. Eleven is the published method, not a cap.
Run a free AI visibility check
BYOK · $12/mo or $49 once · model calls billed by OpenRouter. Free check: watermark stays, detail stops at the first 3 detail rows. No SeenByLLMs account.