Build with ScrapeLLM

Get your first response in under 2 minutes.

# Responses can take up to 300 s — set --max-time accordingly
curl --max-time 300 \
  "https://api.scrapellm.com/scrapers/chatgpt?api_key=YOUR_KEY&prompt=Best+coffee+in+Tokyo%3F&country=JP"

Every request must include your API key via the api_key query parameter or the X-API-Key request header.

# Option A — query parameter
GET /scrapers/chatgpt?api_key=your_api_key&prompt=…

# Option B — request header (preferred in production)
GET /scrapers/chatgpt?prompt=…
X-API-Key: your_api_key

Scrape ChatGPT's real user interface and receive the full reply with inline links and optional markdown token tree.

import requests

r = requests.get(
  "https://api.scrapellm.com/scrapers/chatgpt",
  params={"api_key": "your_key", "prompt": "Best coffee in Tokyo?", "country": "JP"},
  timeout=300,
)
print(r.json()["result"])

const p = new URLSearchParams({api_key:"your_key", prompt:"Best coffee in Tokyo?", country:"JP"});
const r = await fetch(`https://api.scrapellm.com/scrapers/chatgpt?${p}`, {signal:AbortSignal.timeout(300_000)});
const data = await r.json();
console.log(data.result);

curl --max-time 300 \
  "https://api.scrapellm.com/scrapers/chatgpt?api_key=your_key&prompt=Best+coffee+in+Tokyo%3F&country=JP"

{
  "scraper":         "chatgpt",
  "status":          "done",
  "job_id":          "job_abc123",
  "result":          "Tokyo's best coffee…",
  "result_markdown": "**Tokyo's** best coffee…",
  "links": [{"text":"Fuglen Tokyo","url":"https://fuglen.no/…"}],
  "llm_model":       "gpt-4o",
  "credits_used":    3,
  "elapsed_ms":      8421.5,
  "cached":          false
}

ChatGPT-specific parameters

Scrape Perplexity AI and receive the answer with numbered web citations and suggested follow-up questions.

import requests

r = requests.get(
  "https://api.scrapellm.com/scrapers/perplexity",
  params={
    "api_key":    "your_key",
    "prompt":     "Best treatments for insomnia?",
    "country":    "US",
    "web_search": True,
  },
  timeout=300,
)
print(r.json()["sources"])

{
  "scraper":  "perplexity",
  "result":   "CBT-I is considered the gold standard…",
  "sources": [{
    "title":   "Insomnia - Mayo Clinic",
    "url":     "https://mayoclinic.org/…",
    "snippet": "CBT-I involves…"
  }],
  "related_questions": ["What causes insomnia?"],
  "credits_used": 3
}

Perplexity-specific parameters

Scrape Grok (xAI) and receive the answer alongside web sources and X (Twitter) posts — Grok's unique dual-source response format.

⚠ Country codes JP and TW are not supported.

import requests

r = requests.get(
  "https://api.scrapellm.com/scrapers/grok",
  params={
    "api_key": "your_key",
    "prompt":  "Latest news on AI?",
    "country": "US",
    "mode":    "MODEL_MODE_AUTO",
  },
  timeout=300,
)
data = r.json()
print(data["web_sources"], data["x_results"])

{
  "scraper":     "grok",
  "mode":        "MODEL_MODE_AUTO",
  "result":      "OpenAI announced…",
  "web_sources": [{"title":"OpenAI Blog","url":"…"}],
  "x_results": [{
    "user_name": "sama",
    "text":      "Excited about GPT-5…",
    "view_count": 1200000
  }],
  "llm_model": "grok-3",
  "credits_used": 3
}

Grok-specific parameters

Scrape Google Gemini and receive the full markdown answer with rich citation metadata including snippets, highlights, and favicons.

⚠ Country codes JP and TW are not supported.

import requests

r = requests.get(
  "https://api.scrapellm.com/scrapers/gemini",
  params={
    "api_key": "your_key",
    "prompt":  "Explain quantum entanglement.",
    "country": "US",
  },
  timeout=300,
)
print(r.json()["citations"])

{
  "scraper":    "gemini",
  "result":     "Quantum entanglement…",
  "citations": [{
    "title":        "Quantum Entanglement - Wikipedia",
    "url":          "https://en.wikipedia.org/…",
    "snippet":      "When particles become entangled…",
    "website_name": "Wikipedia",
    "highlights":   ["non-locality"]
  }],
  "llm_model":   "gemini-2.0-flash",
  "credits_used": 3
}

No Gemini-specific parameters beyond the common set.

Scrape Microsoft Copilot in one of five modes — from quick chat to deep study guides — and receive citations and all outbound links.

⚠ Country codes JP and TW are not supported.

import requests

r = requests.get(
  "https://api.scrapellm.com/scrapers/copilot",
  params={
    "api_key": "your_key",
    "prompt":  "Best coffee shops in London?",
    "country": "GB",
    "mode":    "search",
  },
  timeout=300,
)
print(r.json()["citations"])

{
  "scraper":    "copilot",
  "mode":       "search",
  "result":     "London's best coffee…",
  "citations": [{
    "title": "Time Out London",
    "url":   "https://timeout.com/…"
  }],
  "links":       ["https://timeout.com/…"],
  "credits_used": 3
}

Copilot-specific parameters

Scrape Google AI Mode and receive the AI answer alongside rich citations and the supporting web search results panel.

⚠ Country codes JP and TW are not supported.

import requests

r = requests.get(
  "https://api.scrapellm.com/scrapers/google_ai_mode",
  params={
    "api_key": "your_key",
    "prompt":  "Best running shoes 2026",
    "country": "US",
  },
  timeout=300,
)
data = r.json()
print(data["citations"], data["search_results"])

{
  "scraper": "google_ai_mode",
  "result":  "The best running shoes of 2026…",
  "citations": [{
    "title":        "Runner's World",
    "url":          "https://runnersworld.com/…",
    "website_name": "Runner's World"
  }],
  "search_results": [{
    "title": "Best Shoes 2026",
    "url":   "https://…"
  }],
  "credits_used": 3
}

No Google AI Mode-specific parameters beyond the common set.

These query parameters are accepted by every scraper endpoint.

Every 200 response includes these fields regardless of which scraper was used.

All errors return a JSON body with a detail field describing the problem.

Use the interactive builder to configure parameters, generate code in Python, JavaScript, or cURL, and test live requests against the API.