ChatGPT Search Scraper: Answers, Citations & Sources API

Give it a list of prompts and get one row per prompt: ChatGPT's answer with web search (Markdown), every source it cites (rank, URL, domain, title), the web search results it read, the searches it ran (fan-out queries), and the brands and products it names.

A ChatGPT scraper alternative: no browser, no proxies, no CAPTCHAs, no OpenAI account. Prompts run five at a time, usually 15 to 30 seconds each; a prompt not answered within 3 minutes gets a free row. You only pay for prompts that got an answer: $0.008 per answer, plus $0.001 per run.

Where the answers come from: this Actor gets ChatGPT's answers (with web search) through a commercial data API that collects them from ChatGPT. Answers vary with the model, the time and the location, so they can differ from what you see in your own ChatGPT account.

Typical uses

• AI search visibility (GEO / AEO): which sites does ChatGPT cite, which brands and products does it recommend, for the prompts your buyers type?
• Citation research: sources and citedDomains show the pages ChatGPT relies on; searchResults shows what it read but did not cite.
• Query research: fanOutQueries shows the web searches ChatGPT ran behind one prompt: keywords to target.
• Answers with sources for your app or agent: ChatGPT search as plain JSON, with the citations kept apart from the text.
• Weekly reports: schedule a run and send the rows to Google Sheets, Looker Studio, Slack, n8n, Make or Zapier.

What you get for each prompt

Field	Description
prompt	The prompt asked
answer, answerLength	ChatGPT's answer as Markdown (headings, lists, tables, images, links)
sources, sourcesCount	Each cited source: rank, url, domain, title, snippet, publicationDate (up to 50)
citedDomains	Domains cited, in order, without duplicates
searchResults, searchResultsCount	Web search results ChatGPT read: rank, url, domain, title, description (up to 50)
fanOutQueries	Web searches ChatGPT ran for the answer
brandEntities	Brands and products named in the answer: name (as written), title, category, urls
products	The brand entities of category product
model, chatgptUrl, answeredAt	Model reported by the data source, a link that opens the prompt in ChatGPT, when it answered
found, status, message	ANSWERED for a billed row; free rows say why: EMPTY_ANSWER, UNREADABLE, REJECTED, PROVIDER_ERROR, NOT_CHECKED
country, languageCode, checkedAt	Location and language of the answer; when the run started

Missing values are null or empty lists, never guessed.

Input

{
  "prompts": ["best standing desk for small spaces", "Which CRM do small agencies recommend in 2026?"],
  "country": "US",
  "languageCode": "en",
  "includeSearchResults": true
}

• prompts: up to 200 prompts per run, 1 to 500 characters each. Each prompt is one answer.
• country (default US): US, GB, CA, AU, IE, IN, DE, AT, CH, FR, BE, ES, MX, IT, NL, PT, BR, PL, SE, JP.
• languageCode (optional): two-letter code; default is the main language of the country.
• includeSearchResults (default true): turn it off to keep only the count of search results. The price does not change.

Pricing

Pay per event, no subscription:

• ChatGPT answer: $0.008 per prompt that got a non-empty answer, at every Apify plan, free plan included.
• Actor start: $0.001 per run.

Empty answers, prompts the data source refused or failed, answers that came back in an unexpected format, and prompts not reached (budget or timeout) get a free row with found: false and the reason. The run never spends more than the maximum cost you set.

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

Good to know

• Answers vary from one run to the next, even for the same prompt: track trends over several runs, not a single answer.
• Not every answer cites sources: ChatGPT searches the web when it decides to, so sources can be empty on general-knowledge prompts (the answer is still billed).
• For one table of brand mentions across ChatGPT, Gemini and Google AI Mode, use AI Visibility Tracker; for Perplexity, Perplexity Search API.
• Not affiliated with OpenAI.

What this is not

• Not the OpenAI API: you cannot pick the model, set a system prompt or continue a conversation. One prompt, one fresh answer.
• Not a login-based scraper: no account, no chat history, no personal context.
• Not a rank tracker for Google: for Google results use a SERP tool.

Run summary

Each run stores a summary in the key-value store (OUTPUT): prompts asked, answers billed, empty answers (promptsEmpty), prompts that failed and are worth a retry (promptsFailed), unreadable answers (promptsUnreadable), 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~chatgpt-search-scraper/run-sync-get-dataset-items?token=YOUR_APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"prompts": ["best standing desk for small spaces"]}'

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

AI agents (MCP tool): add https://mcp.apify.com?tools=plainrows/chatgpt-search-scraper to Claude, Cursor or any MCP client, then ask "What does ChatGPT answer to these prompts, and which sites and products does it cite?".

FAQ

Is this a ChatGPT search API? Yes: one call per list of prompts, JSON rows back with the answer, its sources and the searches behind it, usable from the Apify API, n8n, Make, Zapier or as an MCP tool.

How do I see if ChatGPT cites my site? Look for your domain in citedDomains (and its position in sources.rank); searchResults tells you if ChatGPT read your page without citing it.

Do I pay for empty answers or errors? No. Only prompts with a non-empty answer are charged ($0.008), plus $0.001 per run.

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 Planner Alternative (Bulk Search Volume & CPC) · Google Keyword Planner API (Google Ads Search Volume & CPC) · Bing Keyword Search Volume & CPC · Keyword Ideas Generator · Google Autocomplete & People Also Ask · Google Trends Explorer · Google Trends Keyword Discovery (Rising & Related) · Website Organic Traffic & Top Keywords · Competitor Keywords API (Ranked Keywords) · Bulk Website Traffic Checker · SEO Competitors & Keyword Gap Finder · PPC Competitor Keywords (SpyFu Alternative) · Bulk Domain Authority Checker · Backlink Checker · Backlink API (Referring Domains & Backlinks Data) · Broken Link Checker · SEO Audit API (On-Page SEO Checker) · AI Search Visibility Tracker (ChatGPT, Gemini, AI Mode) · AI Overview Tracker (Google AI Overview Citations) · Perplexity Search API (Answers & Citations) · AI Citations Finder (LLM Mentions in ChatGPT & AI Overviews) · Keyword Rank Checker (Google SERP Position) · Google Search API (SERP Scraper & Web Results) · Bing Search API (Bing SERP & Web Results) · Google Maps Rank Tracker (Geo-Grid Local SEO)
• Leads and companies: Google Maps Scraper · Google Maps Email Extractor · Google Maps Businesses Without Website · Yellow Pages Australia Scraper · Website Technology Checker · Website Email Extractor · Contact Details Scraper (Emails, Phones, Socials) · ATS Jobs Scraper (Workday, Greenhouse, Lever, Ashby) · Workday Jobs Scraper · Greenhouse Jobs Scraper · Lever Jobs Scraper · Ashby Jobs Scraper · Google Jobs Scraper · Reed.co.uk Jobs Scraper (UK) · jobs.ch & jobup.ch Jobs Scraper (Switzerland) · Arbeitsagentur Jobs Scraper (Germany) · Gupy Jobs Scraper (Brazil) · Catho Jobs Scraper (Brazil) · Vagas.com.br Jobs Scraper (Brazil) · Jobs.cz Scraper (Czech Republic) · Remote Jobs Scraper (Himalayas, RemoteOK, Jobicy) · Google Ads Scraper (Transparency Center) · Company Registry Lookup (SIRENE, LEI, Brreg) · Public Tenders Search (EU TED & UK)
• E-commerce and travel: Amazon Product Search Scraper · Amazon Keyword Research (Search Volume & Reverse ASIN) · Google Shopping Prices Scraper · Google Hotels Prices Scraper · Google Hotels Scraper · Tripadvisor Reviews Scraper · Trustpilot Reviews Scraper
• Marketplaces and classifieds: Property Finder Scraper (UAE) · Redfin Scraper (US Homes for Sale) · Otomoto Scraper (Polish Car Listings) · Otodom Scraper (Poland Property) · Imovirtual Scraper (Portugal Property) · Standvirtual Scraper (Portugal Cars) · Immoweb Scraper (Belgium Property) · Blocket Scraper (Sweden) · FINN.no Scraper (Norway) · coches.net Scraper (Spain Cars) · 2dehands Scraper (Belgium) · Storia Scraper (Romania Property) · Autovit Scraper (Romania Cars) · FINN.no Car Scraper (Norway Cars) · DBA Scraper (Denmark) · Tori.fi Scraper (Finland)
• Content and media: Google News Scraper · YouTube Search Scraper · YouTube Search API (JSON) · YouTube Comments Scraper · Google Images Scraper · Website Screenshot API · PDF to Markdown Converter (PDF to Text for AI & RAG) · Google Maps Reviews Scraper · Google Reviews Scraper · App Store & Google Play Reviews Scraper · App Store & Google Play App Details Scraper · Google Play Reviews Scraper · App Store Reviews Scraper

Support

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