search
NetNut Dashboard

Example Requests

Step-by-step examples demonstrating how to send prompts to the ChatGPT Scraper API, including basic requests, Web Search mode, and follow-up prompts.

hashtag
Single Prompt

In this example, we send a simple LLM Scraper request asking for SEO tips for companies. The request enables web_search so the model can enrich its answer with up-to-date information and citations.

Request:

curl -X POST https://llm-scraper.netnut.io/search  \
  -U "username:password" \
  -H "Content-Type: application/json" \
  -d '{
    "engine": "chatgpt",
    "prompt": "what is the best headphones in 2025",
    "web_search": true
  }'

Example Response:

Example shape – your actual traceID, timestamps, and content will vary.

{
  "traceID": "f2c90481-f71d-4896-8588-dcee9ea7f482",
  "model": "gpt-5.2",
  "timestamp": "2025-12-08T11:39:19.269Z",
  "request_duration": 22.29,
  "process_duration": 22.29,

  "response": {
    "prompt": "what is the best headphones in 2025",

    "text_markdown": "Here are some of the **best headphones you can buy in 2025**, based on expert reviews:\n\n### Sony WH-1000XM6\n*$399.99*\n\n### Bose QuietComfort Ultra\n*$399.00*\n\nLet me know if you want recommendations by use case.",

    "text": "Here are some of the best headphones you can buy in 2025, including Sony WH-1000XM6 and Bose QuietComfort Ultra. These models are widely praised for sound quality, comfort, and noise cancellation.",
    
    "model_search_queries": [
            "best headphones 2025 reviews and ranking",
            "top headphones to buy in 2025 best headphones list"
        ],
        
    "widgets": [
      {
        "type": "product_carousel",
        "items": [
          {
            "type": "product_card",
            "name": "Sony WH-1000XM6 Wireless Noise Canceling Headphones",
            "price": 399.99,
            "currency": "$",
            "merchant": "Sony + others"
          },
          {
            "type": "product_card",
            "name": "Bose QuietComfort Ultra Wireless Headphones",
            "price": 399.00,
            "currency": "$",
            "merchant": "Bose + others"
          }
        ]
      }
    ],

    "citations_found": true,
    "citations": [
      {
        "id": "1",
        "title": "The Best Headphones of 2025",
        "url": "https://example.com/best-headphones-2025",
        "section": "citations"
      },
      {
        "id": "2",
        "title": "Top Noise Cancelling Headphones in 2025",
        "url": "https://example.com/noise-cancelling-headphones-2025",
        "section": "more"
      }
    ]
  },

  "raw_response": [
    "data: {\"type\":\"server_ste_metadata\",\"metadata\":{\"model_slug\":\"gpt-5.2\",\"cluster_region\":\"westus3\",\"request_id\":\"c4a36d48-5164-4168-ad18-4bfe82f59e95\"}}",
    "",
    "event: delta",
    "data: {\"v\":{\"metadata\":{\"search_model_queries\":{\"type\":\"search_model_queries\",\"queries\":[\"best headphones 2025 reviews and ranking\",\"top headphones to buy in 2025 best headphones list\"]}}}",
    "",
    "event: delta",
    "data: {\"v\":{\"metadata\":{\"search_result_groups\":[{\"domain\":\"example.com\",\"entries\":[{\"url\":\"https://example.com/article\",\"title\":\"Best headphones to buy in 2025\"}]}]}}}"
  ]
}

hashtag
In this response:

  • model specifies the AI model used to generate the response (for example, gpt-5.2).

  • text_markdown contains the full markdown-formatted answer, including headings, lists, and emphasis.

  • text contains a plain-text version of the same content, with all markdown removed.

  • model_search_queries contains an array of search queries generated by the model during the request, reflecting how the original prompt was expanded or refined when performing web search or information retrieval.

  • widgets contains structured UI components extracted from the model output (such as product carousels and product cards).

  • citations_found indicates whether any sources were detected.

  • citations is a flat list of source objects, each tagged with:

    • section: "citations" for primary sources

    • section: "more" for additional references

  • raw_response contains the raw streamed model output, exactly as captured from the session before parsing. This includes server events, metadata, search fan-out queries, and intermediate results.

hashtag
Request with Additional Follow-Up Prompt

In this example, we send:

  • An initial prompt: best SEO tips for new companies.

  • An additional follow-up prompt: “and what about ecommerce companies?”.

The LLM Scraper runs both prompts within the same flow and returns:

  • response – main answer for the original prompt.

  • follow_up_response – answer to the follow-up prompt.

Request (Prompt + Additional Prompt)

Example Response (Prompt + Additional Prompt)

This is a real example payload from the system; line breaks in text are preserved as returned by the LLM.

In this response:

  • response holds the answer to the main prompt (“what is the best seo tips for new companies”).

  • follow_up_response holds the answer to the additional prompt (“and what about ecommerce companies?”).

  • Each block has its own:

    • prompt

    • text (markdown content)

    • citations_found

    • citations array (which may be empty for some follow-ups, as in this example).

This pattern is ideal for SEO customers who:

  • Ask a general question first (e.g., overall SEO tips).

  • Then add a more specific follow-up (e.g., ecommerce-specific tactics) and want both results in a single, structured response.

PreviousResponse FieldsNextGoogle AI Mode

Last updated