AI Search Visibility Tracker: ChatGPT, Gemini & Google AI Mode

Ask ChatGPT, Gemini and Google AI Mode the questions your customers ask, and see in one table whether your brand is mentioned, whether your website is cited, which competitors show up, and which sites the answers rely on. Each answer comes with its sources, the web searches the engine ran (ChatGPT) and, if you want, the full text.

It calls a commercial data API instead of driving browsers or accounts: no proxies, no logins, no CAPTCHAs. Each answer takes about 20 seconds; questions run three at a time. You only pay for answers returned: engines that fail, empty answers and questions nobody answered are free.

Typical uses

• AI search visibility (GEO / AEO): are you recommended when people ask "best CRM for a small agency" or "is Acme reliable?" — and in which position compared with your competitors?
• Citation tracking: which pages do the engines cite for your topics, and is your site one of them? Spot the reviews and comparison sites you need to be on.
• Competitor monitoring: who gets named instead of you, engine by engine, country by country.
• Content planning: the fan-out queries show what ChatGPT searched for to answer; the cited pages show what content wins.
• Weekly reports and alerts: schedule a run and send the table to Google Sheets, Looker Studio, Slack, n8n, Make or Zapier.

What you get for each answer

One row per question and engine:

Field	Description
prompt, engine	Your question; chatgpt, gemini or google_ai_mode
brandMentioned, brandMentionRank	Your brand appears in the answer; 1 = it is named before your competitors
brandDomainCited, brandDomainCitationRank	Your website is among the sources or links, and at which position
competitorsMentioned, competitorDomainsCited	Competitors named (in order of appearance) and competitor sites cited
citedDomains, sources	Domains and sources (title, domain, URL) the answer cites, in order, up to 20
fanOutQueries	Web searches ChatGPT ran to answer
answerText, answerLength	The answer as Markdown (optional) and its length
model	Model that answered, when the engine reports it
country, languageCode, checkedAt	Where and when the question was asked

Brand and competitor names are matched as whole words, in any case ("Acme" does not match "Acmex"). Images in an answer are not counted as citations. Missing values are null, never guessed.

In the run's Output tab, the Cited sources view lists one row per source (question, engine, domain, title, URL), ready to export to CSV or Google Sheets: see which pages the AI engines rely on for your questions. From the API: .../items?view=sources.

Input

{
  "prompts": ["best standing desk for small spaces", "best electric standing desk under $500"],
  "brandNames": ["FlexiSpot"],
  "brandDomain": "flexispot.com",
  "competitors": ["Uplift", "Desky", "IKEA"],
  "competitorDomains": ["upliftdesk.com"],
  "engines": ["chatgpt", "gemini", "google_ai_mode"],
  "country": "US"
}

• prompts: up to 200 questions per run, kept as you type them. Each one is asked once to each engine.
• brandNames (optional): your brand and its variants. Without it you still get the answers, sources and competitors.
• brandDomain (optional): your site; subdomains count, a path is ignored (acme.com/fr = acme.com). Gemini's sources are mostly Google Shopping and Google Search pages, so Gemini rarely shows your own site as cited.
• competitors, competitorDomains (optional): up to 10 each.
• engines: any of chatgpt (ChatGPT with web search), gemini, google_ai_mode. Default: all three.
• country: US, GB, CA, AU, IE, IN, DE, AT, CH, FR, BE, ES, MX, IT, NL, PT, BR, PL, SE, JP. languageCode defaults to the country's main language.
• includeAnswerText (default true): turn off for lighter rows.

Output example

Real row from 2026-10-09 (ChatGPT, US; answer text and source links shortened here):

{
  "prompt": "best standing desk for small spaces",
  "engine": "chatgpt",
  "found": true,
  "model": "gpt-6",
  "brandMentioned": true,
  "brandMentionRank": 1,
  "brandDomainCited": false,
  "brandDomainCitationRank": null,
  "competitorsMentioned": [
    "Desky",
    "Uplift"
  ],
  "competitorDomainsCited": [
    "upliftdesk.com"
  ],
  "citedDomains": [
    "staples.com",
    "desky.com",
    "bestbuy.com",
    "wfhworld.com",
    "deskfinds.com",
    "upliftdesk.com"
  ],
  "sources": [
    {
      "title": "Shop at Staples",
      "domain": "staples.com",
      "url": "https://www.staples.com/flexispot-ec2-40-w-electric-rectangular-adjustable-standing-desk-white-ec2w-4024w/product_24686947"
    },
    {
      "title": "Shop at Desky",
      "domain": "desky.com",
      "url": "https://desky.com/products/dual-small-standing-desk"
    }
  ],
  "fanOutQueries": [
    "best compact standing desks for small spaces 2026 reviews dimensions"
  ],
  "answerText": "For a small bedroom, apartment, or home-office corner, I’d look for a standing desk that’s **40–42 inches wide, around 24 inches deep, and electrically height-a…",
  "answerLength": 4846,
  "country": "US",
  "languageCode": "en",
  "checkedAt": "2026-10-09T13:21:40.000Z",
  "schemaVersion": "1.0"
}

Pricing

Two charges, no subscription, no fee per run (see the Pricing tab for the current prices):

• AI answer: per answer returned (one engine's answer to one question). Three engines and ten questions = up to 30 answers.
• Question asked: a small fixed fee once per question that got at least one answer, whatever the number of engines. Each engine asked costs us the same whether it answers or comes back empty; this fee pays for the empty ones, so empty answers themselves are never billed.

Engines that fail, empty answers and questions with no answer at all are never billed (no question fee either). The run never spends more than the maximum cost you set: it asks only the engines the remaining budget can pay for.

Free Apify plan: each run asks the first 3 questions, so you can test the output. A paid Apify plan removes this limit.

Good to know

• AI answers change from one run to the next, even for the same question: track trends over several runs (a weekly schedule), not a single answer.
• Answers are those of a logged-out user in the chosen country and language, not of a personal account with its history.
• In our test on 2026-10-09 (one shopping question), ChatGPT cited 6 sources, Google AI Mode 3, and Gemini's 5 sources were Google Shopping product pages: what counts as a citation differs from one engine to another.
• Not affiliated with OpenAI or Google.

Run summary

Each run stores a summary in the key-value store (OUTPUT): questions run, questions with no answer, answers billed, answers that came back without text (not billed), engines that failed, and anything skipped because of the free-plan or budget limit.

Use it from code, automations or AI agents

API (one call, results returned directly):

curl -X POST "https://api.apify.com/v2/acts/plainrows~ai-search-brand-visibility/run-sync-get-dataset-items?token=YOUR_APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"prompts": ["best CRM for a small agency"], "brandNames": ["Acme"], "brandDomain": "acme.com"}'

n8n, Make, Zapier: use the Apify integration, pick plainrows/ai-search-brand-visibility, schedule it weekly and send the rows to Google Sheets or Slack.

AI agents (MCP): add https://mcp.apify.com?tools=plainrows/ai-search-brand-visibility to Claude, Cursor or any MCP client.

FAQ

How do I check if ChatGPT recommends my brand? Put the questions your customers ask in prompts, your brand in brandNames, and run. brandMentioned and brandMentionRank tell you, engine by engine.

Does it use my ChatGPT account? No account is needed or used. Answers come from a commercial data API.

Which AI engines? ChatGPT (with web search), Gemini and Google AI Mode. Perplexity and Claude are not included.

Why do two runs give different answers? AI engines vary their answers. Compare several runs over time.

How do I also get how often a keyword is used in AI tools? Keyword Search Volume adds an AI search volume per keyword (from October 24, 2026).

More PlainRows tools

Same pay-per-result pricing, same clean JSON, built to chain together in n8n, Make, Zapier or AI agents.

• SEO and keywords: Keyword Search Volume, CPC & Difficulty · Keyword Ideas Generator · Google Autocomplete & People Also Ask · Google Trends Explorer · Website Organic Traffic & Top Keywords · Bulk Domain Authority Checker · Backlink Checker · Broken Link Checker
• Leads and companies: Google Maps Scraper · Website Technology Checker · ATS Jobs Scraper (Greenhouse, Lever, Ashby, SmartRecruiters) · Google Jobs Scraper · Google Ads Transparency Scraper · Company Registry Lookup (SIRENE, LEI, Brreg) · Public Tenders Search (EU TED & UK)
• E-commerce and travel: Amazon Product Search Scraper · Google Shopping Prices Scraper · Google Hotels Prices Scraper
• Content and media: Google News Scraper · YouTube Search Scraper · Google Images Scraper · Website Screenshot API · App Reviews Scraper (Google Play & App Store)

Compared with other APIs: Is your brand in ChatGPT, Gemini and Google AI Mode answers?

Support

Missing a field or found a bug? Open an issue on this page. Issues are read and answered.