Scrape Google Gemini responses via API. Returns parsed JSON with full text and markdown, grounded sources with confidence scores, citation pills, and Gemini's internal reasoning steps. Python, cURL, and Node.js examples below.
Built for developers doing AI brand monitoring on Gemini, source-attribution research, competitive intelligence across Google's AI stack, and content generation at scale — without managing CAPTCHAs, rotating proxies, session state, or Google's anti-bot defenses.
-
Get an API key at cloro.dev.
-
Send a request:
curl -X POST https://api.cloro.dev/v1/monitor/gemini \ -H "Authorization: Bearer YOUR_API_KEY" \ -H "Content-Type: application/json" \ -d '{"prompt": "What are the top climate risks for coastal real estate?"}'
-
Parse the returned JSON —
result.text,result.markdown,result.citationPills[],result.sources[].
Full examples in Python, cURL, and Node.js below.
The Gemini scraper handles the rendering, parsing, and delivery of results in your requested format. You provide your prompt, API credentials, and optional parameters as shown below.
import json
import requests
# API parameters
payload = {
'prompt': 'Explain the concept of quantum entanglement',
'country': 'US',
'include': {
'markdown': True
}
}
# Get a response
response = requests.post(
'https://api.cloro.dev/v1/monitor/gemini',
headers={'Authorization': 'Bearer YOUR_API_KEY'},
json=payload
)
# Print response to stdout
print(response.json())
# Save response to a JSON file
with open('response.json', 'w') as file:
json.dump(response.json(), file, indent=2)curl -X POST https://api.cloro.dev/v1/monitor/gemini \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"prompt": "Explain the concept of quantum entanglement",
"country": "US",
"include": {
"markdown": true
}
}'const axios = require("axios");
const payload = {
prompt: "Explain the concept of quantum entanglement",
country: "US",
include: {
markdown: true,
},
};
axios
.post("https://api.cloro.dev/v1/monitor/gemini", payload, {
headers: {
Authorization: "Bearer YOUR_API_KEY",
"Content-Type": "application/json",
},
})
.then((response) => {
console.log(response.data);
})
.catch((error) => {
console.error("Error:", error);
});| Parameter | Description | Default value |
|---|---|---|
prompt* |
The prompt or question to send to Gemini (1-10,000 characters) | – |
country |
Optional country/region code for localized results (e.g., US, JP) |
US |
state |
Optional US state code for state-level geo-targeting (e.g., CA, TX, NY). Only valid when country is US. See supported codes. |
– |
include.markdown |
Include response in Markdown format when set to true | false |
include.html |
Include URL to full HTML response when set to true (URL expires after 24h) | false |
include.rawResponse |
Include raw streaming response events for debugging | false |
- Mandatory parameters ** Gemini is currently not available in European countries (EEA/EU).
The Gemini Scraper API returns a structured JSON object containing Gemini's AI-generated response and metadata.
Structured JSON output snippet:
{
"success": true,
"result": {
"text": "Quantum entanglement is a physical phenomenon that occurs when a group of particles are generated, interact, or share spatial proximity in a way such that the quantum state of each particle of the group cannot be described independently of the state of the others...",
"sources": [
{
"position": 1,
"url": "https://en.wikipedia.org/wiki/Quantum_entanglement",
"label": "Wikipedia",
"description": "Quantum entanglement is a physical phenomenon..."
}
],
"html": "https://storage.cloro.dev/results/c45a5081-808d-4ed3-9c86-e4baf16c8ab8/page-1.html", // URL expires after 24 hours
"markdown": "**Quantum entanglement** is a physical phenomenon..."
}
}Gemini responses include a sources array with citation details:
| Field | Type | Description |
|---|---|---|
position |
integer | Position order of the source in the response |
url |
string | Direct URL to the source content |
label |
string | Source name or publication |
description |
string | Brief description of what the source contains |
When the Gemini answer carries pill chips (including section-summary chips), the result.citationPills array exposes each cited source as a self-contained entry. When a chip cites N sources, the array contains N entries sharing the same citationPillId but with per-source label, url, and domain. Group by citationPillId to recover pill-level structure. The field is omitted when no pills are present.
| Field | Type | Description |
|---|---|---|
label |
string | Per-source title from the sources rail (e.g. "Quantum computing — Wikipedia"). Always present; may be an empty string when the rail has no title for this source — read domain / url for source identity in that case. |
citationPillId |
integer | 1-based ordinal shared by all entries from the same chip. |
url |
string | Direct URL of the cited source. |
domain |
string | Host extracted from url, for grouping and display. |
description |
string | Source snippet when present. Omitted when absent. |
position |
integer | 1-based position of this source in the sibling result.sources array. |
- Reasoning: Solve logic puzzles, math problems, and analytical tasks.
- Content creation: Generate blog posts, marketing copy, and creative stories.
- Code generation: Write, debug, and explain code snippets in various languages.
- Data analysis: Analyze and summarize datasets or documents.
- Educational tutoring: Create learning materials and explanations.
- Research assistance: Synthesize information from multiple cited sources.
- Simple integration: Clean API design with documentation and examples.
- Reliable performance: >99% uptime and low latencies.
- No infrastructure hassle: We handle rate limiting and browser management.
- Gemini access: Access to Google's Gemini models.
- Developer support: Responsive support team to help with integration and troubleshooting.
Any website is legal to be scraped as long as the information is publicly accessible.
cloro's Gemini endpoint provides access to Google's Gemini AI with:
- Structured source citations with URLs, labels, and descriptions
- Structured data extraction for direct integration
- Low-latency infrastructure with high availability
We don't recommend putting any timeout, given that our system retries automatically. We recommend setting up a retry mechanism in case of failure.
Yes, you can specify country codes. Note that Gemini is currently unavailable in European countries (EEA/EU).
Gemini excels at reasoning, coding, creative writing, and complex analytical tasks.
For detailed documentation, advanced features, and integration guides, visit:
- API documentation: cloro.dev/docs
- Gemini scraper page: cloro.dev/gemini
- AI Mode - Extracts structured data from Google AI Mode for general knowledge queries, workflow optimization, and technical guidance.
- AI Overview - Extracts structured data from Google AI Overview for comprehensive search result analysis and AI-curated insights.
- ChatGPT - Extracts structured data from ChatGPT with advanced features including shopping cards, raw response data, and query fan-out.
- Copilot - Extracts structured data from Microsoft Copilot for development tools, Microsoft ecosystem research, and enterprise-focused queries.
- Gemini - Extracts structured data from Google Gemini for complex reasoning, content generation, and cited research.
- Google Search - Extracts structured data from Google Search results, including organic results, People Also Ask questions, related searches, and optional AI Overview data.
- Google News - Extracts structured news articles from Google News with titles, snippets, sources, dates, and thumbnail images for news monitoring and media tracking.
- Grok - Extracts structured data from Grok for current events, news tracking, and real-time information gathering.
- Perplexity - Extracts comprehensive structured data from Perplexity AI with real-time web sources, automatically detecting and extracting rich data objects.
If you have questions or need support, join our community at r/cloroapi.
Built with ❤️ by the cloro team