AI SEARCH INSIGHTS

ChatGPT Scraper API

Turn AI answers into your next move.

Collect ChatGPT answers, cited sources and retrieved sources through one API. Turn AI search responses into data for brand visibility research, citation analysis and your GEO workflows.

Try ChatGPT Scraper View API example

Answers & sources
Residential proxies
JSON

AI Search Insights

YOUR PROMPTGLOBAL

Which tools help teams research AI search visibility?

ChatGPT

CAPTURED DATAEXAMPLE

result_text

Evaluate tools by their answer capture, source coverage and regional controls. Use that evidence in your own workflow to study brand mentions.

Read the answer alongside its sources

Illustrative output · field examples, not a live capture

01 / THE OVERVIEW

What is a ChatGPT scraper?

AdsCrawl ChatGPT Scraper is the ChatGPT capture capability in AI Search Insights. Send a prompt, choose a proxy region and browser language, then inspect the answer and available source and ad information. Use the web console to explore a request, or call POST /ai-search to bring the captured data into your own application.

02 / THE DATA

The answer. Its sources. The context.

Keep traceable data for content research, brand visibility and GEO analysis. Cited and retrieved sources are returned separately for closer review.

01Answer textresult_text: The captured ChatGPT answer, ready for reading, text processing or your own brand-mention analysis.
02Cited sourcescited_sources: Sources cited in the answer, when available. Inspect the references that support the response.
03Retrieved sourcesretrieved_sources: Source information retrieved during the response, when available. Keep it separate from the sources actually cited in the answer.
04Ad informationads: Advertising information observed during capture, when present. An answer does not necessarily contain ads.
05Proxy regioncountry_code: The proxy region associated with the capture. Use it alongside the observed exit network when comparing regional results.
06Browser languagelocale: The browser-language context for the capture. It is distinct from the language you request in your prompt.
07Observed exit networkcapture_context.network: The observed network context of the capture, useful for checking the actual exit against the selected proxy region.
08Task IDtask_id: The identifier for the capture task. Keep it with your results to trace a request when investigating an issue.

Citations, retrieved sources and ads depend on what appears in each capture and may be empty.

03 / THE CONTEXT

Put every prompt in context.

Choose a proxy region, browser language and timeout in the console, then carry those settings into your API request.

Dynamic residential proxies

Each capture request is routed through a dynamic residential proxy. Run a browser-based capture without configuring a separate proxy for this request.

Selected region, observed exit

Choose a supported proxy region, or use Global for a random exit across supported regions. Check capture_context.network for the observed exit; a region selection alone does not guarantee access or a particular answer.

Browser language and answer language

Use Automatic to follow the proxy region, or choose a browser language. To request an answer in a particular language, say so in the prompt.

A timeout that fits the request

Set the overall timeout from 10 to 120 seconds in the console. API requests express it in milliseconds as timeoutMs; keep your client's timeout longer than the capture timeout.

04 / FROM DATA TO DECISIONS

See how AI talks about your market.

Generative engine optimization (GEO) starts with observing real answers. Bring captures into your own analysis to find content opportunities worth testing.

Research brand mentions

Capture answers to the questions your audience asks, then examine whether and how your brand or products appear. Feed the text into your own AI search visibility or GEO analysis.

Understand cited content

Review the sources cited in an answer and compare them with the retrieved sources. Use the evidence to investigate which pages inform a response and guide further content research.

Compare regional responses

Repeat a question with different supported proxy-region settings and keep the network and language context with each result. Compare wording, mentions and sources in your own workflow.

05 / API INTEGRATION

Try it in the console. Ship it in your code.

One /ai-search endpoint, starting with a single prompt. Feed structured results into your research tools and data workflows.

1
Configure a capture
Open AI Search Insights, select ChatGPT and enter a prompt. Choose the proxy region, browser language and timeout for the question you want to explore.
2
Inspect the response
Run the request in the console and review the answer, available sources and capture context. Check what was actually returned before using it in your analysis.
3
Bring it into your application
Open API integration and copy the request that follows your current settings. Replace YOUR_API_KEY on your server and call POST /ai-search from your own workflow.

Open AI Search Insights

POST /ai-search

curl --request POST 'https://api.adscrawl.net/ai-search' \
  --header 'x-api-key: YOUR_API_KEY' \
  --header 'Content-Type: application/json' \
  --max-time 140 \
  --data '{
  "platform": "chatgpt",
  "query": "Recommend a few good-value phones and include reference sources.",
  "country_code": "GLOBAL",
  "timeoutMs": 120000
}'

Replace YOUR_API_KEY with your API key and run the request on your server. The example sets a 120-second task timeout and allows 140 seconds for the cURL response.

06 / FAQ

Questions about scraping ChatGPT

What does the ChatGPT Scraper API collect?

It captures ChatGPT answer text, cited and retrieved sources, and ad information when those elements are available. The response also exposes proxy-region and browser-language context, the observed exit network and a task identifier.

Does every answer include citations or ads?

No. The content depends on the prompt and the response shown during capture. Some answers have no cited sources or advertising. Treat these fields as observed information, not a promise that every element will appear.

How do I request an answer in another language?

Include the desired language in your prompt, such as asking ChatGPT to answer in German. The browser-language setting controls the browser context and does not itself require ChatGPT to reply in that language.

Does AdsCrawl calculate a brand visibility score?

The scraper provides the answer and source data for your analysis. Define the brand mentions, citation measures and comparisons that matter to your business, then use the captured data in your own scoring or monitoring workflow.

Can I try a request before integrating the API?

Yes. AI Search Insights lets you configure and run a request in the web console. Its API integration panel shows the endpoint and a cURL example based on your settings, so you can move the same configuration into your server-side application.

Which other AI scrapers are available?

ChatGPT is currently available in AI Search Insights. Gemini, Google AI Mode and Google AI Overview are marked Coming soon and cannot yet be selected for a capture.

START WITH ONE QUESTION

Your next insight is in the answer.

Open AI Search Insights to capture an answer and the sources behind it.

Start collecting