AI Rank Tracker - Track Brand Positions in AI Answers

AI Rank Tracker checks where a brand, product, company, tool, or domain appears inside AI answer rankings. It submits bounded ranking prompts to Perplexity, ChatGPT, Gemini, and optional Claude, then returns one record per prompt x platform with target rank, top ranked entities, competitor ranks, citations, source domains, and an AI rank score.

Use it when you need dated evidence for GEO, AEO, and AI-search reporting: "Where does my brand rank when buyers ask AI for the best tools in my category?" Do not use it for classic Google SERP rank tracking, raw Google AI Overview extraction, or website crawlability audits.

Output Fields

Field	Description
platform	AI answer engine checked: Perplexity, ChatGPT, Gemini, or Claude
query	Exact ranking prompt submitted to the AI platform
targetRank	Detected target position in the AI answer shortlist; null when not ranked
rankBucket	rank_1, top_3, top_10, mentioned_unranked, or not_found
aiRankScore	0-100 score derived from rank, share of voice, and owned-domain citation
rankedEntities	Top entities detected from the answer's ranked list or fallback mention order
winningEntity	First-ranked entity detected in the answer
competitorRanks	Rank positions for the competitors you supplied
targetShareOfVoice	Target mentions divided by target plus competitor mentions
domainCited	Whether the target domain was cited as a source
citedUrls	Source URLs extracted from citations, annotations, grounding, or links
sourceDomains	Deduplicated domains cited by the AI response
aiAnswerExcerpt	Concise or detailed answer excerpt for review

Pricing

This actor uses Pay per event + usage. Event charges are $0.10 per successful AI rank check, where one check is one prompt on one AI platform. Apify platform usage is charged separately by Apify.

Example run	Event charges
1 prompt x 1 platform	$0.10
3 prompts x 3 default platforms	$0.90
5 prompts x 4 platforms including Claude	$2.00

The actor prints the maximum event-charge cap before it calls any AI provider. Failed provider calls are not billed as rank rows.

Input

{
  "targetName": "Apify",
  "targetDomain": "apify.com",
  "industry": "web scraping tools",
  "competitors": ["Bright Data", "Octoparse"],
  "queryTemplates": ["recommendations", "alternatives", "category_leaders"],
  "platforms": ["perplexity", "chatgpt", "gemini"],
  "maxQueries": 3,
  "maxRankedEntities": 10,
  "responseFormat": "concise"
}

targetName is required. targetDomain is optional but recommended because it enables owned-domain citation tracking. Claude is available when selected explicitly; it is not part of the default platform set to keep typical runs under agent-payment budgets.

Output Example

{
  "platform": "perplexity",
  "query": "Rank the best web scraping tools for agencies.",
  "targetName": "Apify",
  "targetDomain": "apify.com",
  "targetMentioned": true,
  "targetMentionCount": 2,
  "targetMentionPosition": 1,
  "targetRank": 1,
  "rankBucket": "rank_1",
  "aiRankScore": 96,
  "targetShareOfVoice": 0.5,
  "sentiment": "positive",
  "domainCited": true,
  "citedUrl": "https://apify.com/store",
  "rankedEntities": ["Apify", "Bright Data", "Octoparse"],
  "rankedEntityCount": 3,
  "winningEntity": "Apify",
  "competitorRanks": [{"entity": "Bright Data", "rank": 2}],
  "sourceDomains": ["apify.com"],
  "modelUsed": "perplexity/sonar",
  "checkedAt": "2026-07-14T18:00:00.000Z"
}

Compact OUTPUT and detailed RUN_SUMMARY records are also written to the key-value store. The summary includes ranked rate, average target rank, AI rank score, platform scores, top source domains, top competitors, provider diagnostics, event charges, and recommendations.

API

curl -X POST "https://api.apify.com/v2/acts/khadinakbar~ai-rank-tracker/runs" \
  -H "Authorization: Bearer YOUR_APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "targetName": "Apify",
    "targetDomain": "apify.com",
    "industry": "web scraping tools",
    "competitors": ["Bright Data", "Octoparse"],
    "platforms": ["perplexity", "chatgpt", "gemini"],
    "maxQueries": 3
  }'

MCP And Agent Use

Through the Apify MCP Server, this actor is useful for questions like:

• "Check where Apify ranks in AI answers for web scraping tool prompts."
• "Compare my AI rank against these competitors across ChatGPT, Perplexity, and Gemini."
• "Which sources do AI answers cite when ranking my category?"

The actor returns a flat, schema-declared dataset so agents can page results, filter by platform, compare rank buckets, and summarize the RUN_SUMMARY KV record.

Limits And Reliability Notes

AI answer engines change outputs frequently. Treat each run as a dated sample, not a permanent ranking guarantee. The actor records the model used and checkedAt timestamp so recurring runs can be compared over time.

Rank extraction prefers explicit numbered lists, bullet lists, headings, and simple tables. When an answer is prose instead of a clean shortlist, the actor falls back to mention order among the target and supplied competitors. Rows expose both targetRank and rankedEntities so you can inspect the evidence.

Provider calls require owner-managed OPENROUTER_API_KEY; native Gemini uses GEMINI_API_KEY when available and falls back to OpenRouter Gemini online when possible. If a required provider is unavailable and no useful rows are produced, the actor fails honestly with UPSTREAM_FAILED or CONFIG_ERROR and writes OUTPUT plus RUN_SUMMARY.

Related Actors

Use AI Search Brand Monitor when you want the older brand-monitoring contract focused on mention trends and GEO recommendations.

Use AI Search Visibility Tracker when the job is domain citation tracking for topic keywords and the brand name itself is secondary.

Use AI Overviews Tracker when the target is Google AI Overview-specific visibility rather than multi-engine AI answers.

Legal And Data Use

This actor queries AI provider APIs and records generated answer excerpts, citation URLs, and derived rank metrics. Review provider terms and your organization's policies before using outputs in public reports. Do not submit private customer data, secrets, or personal information in prompts.