AI Visibility Tracker (ChatGPT, Perplexity, Gemini)
Pricing
Pay per usage
AI Visibility Tracker (ChatGPT, Perplexity, Gemini)
Track whether ChatGPT, Perplexity and Gemini mention your brand and cite your domain for the questions your customers ask. Uses official APIs with your own keys.
AI Visibility Tracker — are ChatGPT, Perplexity and Gemini recommending you?
Ask the AI engines the questions your customers ask, and get one row per question and engine telling you:
- mentioned — is your brand named in the answer?
- cited — is your domain among the sources the engine cites? (citationRank = its position in the source list)
- competitorsMentioned — which competitors the answer names instead
- citations — every cited source, so you can see who the engine trusts for that question
- answer — the answer text (first 4,000 characters)
Run it on a schedule and you have an AI share-of-voice history for your brand (GEO / AEO / LLM SEO tracking).
How it works
The actor calls each engine's official API with live web search — Gemini with Google Search grounding, Perplexity sonar, and OpenAI Responses API with web search — using your own API key. No scraping, no proxies, no logged-in sessions, so runs are stable and you only pay the engine for what you ask. Add one, two or all three keys; engines without a key are skipped. Keys are secret inputs and are never written to the dataset or log.
Answers from the API can differ from what a logged-in user sees in the consumer app (personalisation, location, model tier). Treat the result as a consistent benchmark over time, not a screenshot of one user's chat.
Input
{"queries": ["best project management tool for small teams", "asana alternatives"],"brand": ["Asana"],"domain": "asana.com","competitors": ["Trello", "Monday", "ClickUp"],"geminiApiKey": "…", "perplexityApiKey": "…", "openaiApiKey": "…"}
Up to 100 questions per run. Any language works — ask in the language your customers use.
Output (one row per question × engine)
query, engine (gemini | perplexity | chatgpt), mentioned, cited, citationRank, competitorsMentioned, citations, answer, checkedAt. If an engine call fails (bad key, quota), the row carries error and the run continues.
Related
- AI Search Readiness Checker (azet-io/ai-search-readiness) — can AI crawlers read your site at all?
- llms.txt Generator (azet-io/llms-txt-generator) — draft an
/llms.txtfrom your sitemap.
Pricing
Free for now (you pay only your own engine API usage). Pay-per-event pricing is planned; it will be announced on this page before it starts.
Made by AZET (azet.io).