As AI search replaces traditional search engines, more user queries, content, and decision-making happen inside models such as ChatGPT, Perplexity, Copilot, Gemini, and Google AI Overviews.
Brands and teams need a way to collect, analyze, and monitor real-time insights from these AI engines—including prompts, answers, citations, rankings, trends, and competitor mentions.

The LLM Chat Scraper API is built for exactly this purpose.

It provides a unified scraping interface to extract structured, real-time data from all major AI models—allowing you to use the results for GEO (Generative Engine Optimization), competitor monitoring, content strategy optimization, and search intelligence.

Getting Started

Using the LLM Chat Scraper API consists of two simple steps:

Step 1: Create a Task

Send a POST request to create a scraping task.
If webhook.url is specified, the result will be automatically pushed when the task completes.

Request Example

curl '{api_host}/api/v2/scraper/request' \
--header 'Content-Type: application/json' \
--header 'x-api-token: {you_api_key}' \
--data '{
  "actor": "scraper.chatgpt",
  "input": {
    "prompt": "Most reliable proxy service for data extraction",
    "country": "US",
    "web_search": true
  },
  "webhook": {
    "url": "http://www.youwebhook.com"
  }
}'

Step 2: Retrieve the Result

Results are stored for 5 minutes. Make sure to fetch them promptly.

Request Example

curl --request GET '{api_host}/api/v2/scraper/result/{task_id}' \
--header 'Content-Type: application/json' \
--header 'x-api-token: {you_api_key}'

Common Parameters

Result Data Structure

Webhook Push Format

If webhook.url is specified, the API sends the result via POST.

HTTP Status Codes

Scrapers Overview

Below are the supported AI model scrapers and their data formats.

1. ChatGPT Scraper

Body Parameters

Response Fields

2. Perplexity Scraper

Key Response Fields

• prompt
• result_text
• related_prompt (related questions)
• web_results (title, URL, snippet)
• media_items (videos, maps, images)
• locations (lat/lng, description, categories, address)

Supports rich structured data for travel, local info, news, and trending topics.

3. Copilot Scraper

Supports multiple modes:
search, smart, chat, reasoning, study

Body Parameters

Response Fields

• result_text
• prompt
• mode
• links
• citations

4. Gemini Scraper

Response Fields

• result_text
• prompt
• citations (favicon, highlights, snippet, website_name)

Supports rich citation structures similar to Google Gemini responses.

5. Google AI Mode Scraper

Used for scraping Google AI Overviews / AIO responses.

Response Fields

Help & FAQ

Billing

If the result is generated but not retrieved within 5 minutes, the request is still billed.
To avoid waste:

• Retrieve results immediately, or
• Configure a webhook to auto-receive results

Data Source

We only scrape public, login-free accessible data, ensuring compliance and privacy protection.

Supported Countries / Regions

(Partial list below)

Full list with 195+ countries is available on request.

Conclusion

The LLM Chat Scraper API gives teams the ability to:

• Monitor brand mentions across all AI chat platforms
• Track competitor presence and ranking in AI answers
• Analyze model outputs, citations, and trends
• Build GEO (Generative Engine Optimization) strategies
• Automate real-time intelligence pipelines
• Access structured data from the entire AI search ecosystem

It is more than a scraper—it's a data infrastructure layer for the AI Search Era.

Contact us to unlock the full GEO data solution —
so every piece of content is backed by data, aligned with algorithm behavior, and positioned for measurable growth.