Grok's citations carry the richest per-source metadata of any AI engine: site name, creator, preview text, favicon, and image — across both the web and X posts. Plus the search queries Grok ran to build its answer, all structured JSON.
4.8 · 33 reviews
Sources and query fan-out are parsed from the real grok.com surface — including references to X posts that only exist there.
import requests
response = requests.post(
"https://api.cloro.dev/v1/monitor/grok",
headers={
"Authorization": "Bearer sk_live_your_api_key_here",
"Content-Type": "application/json"
},
json={
"prompt": "which ev has the best real-world range",
"country": "US",
"include": {
"markdown": True
}
}
)
print(response.json()){
"success": true,
"result": {
"text": "Based on independent range tests, the EVs with the best real-world range are...",
"sources": [
{
"position": 1,
"url": "https://example.com/ev-range-test",
"label": "Real-World EV Range Test Results",
"description": "70-mph highway range tests across 40 EVs",
"preview": "We drove every major EV at a steady 70 mph until empty...",
"siteName": "Example Motors",
"creator": "Jane Doe",
"favicon": "https://example.com/favicon.ico"
}
],
"searchQueries": [
"ev real world range test 2026",
"longest range electric car highway test"
],
"model": "grok-4"
}
} | Field | Type | Description |
|---|---|---|
| position | number | Position index of the source in the response. |
| url / label / description | string | Cited URL, display label, and source description. |
| preview | string | Preview text snippet from the source. |
| siteName / metadataTitle | string | Website name and source metadata title. |
| creator | string | Content creator or author — includes X handles for cited posts. |
| image / favicon | string | Source image and favicon URLs. |
| searchQueries | string[] | Search queries Grok emitted while answering (query fan-out). |
| model | string | Model identifier that produced the answer. |
Pick a plan that fits your volume. Price per credit drops as you scale.
Increased concurrency, overages on credits and credit discounts for annual contracts.
Know moreCredit cost per request varies by provider. The rates below apply to async/batch requests; sync requests add a +2 credit surcharge.
ChatGPT full response includes query fan-out, ads, and shopping data. Google News uses the same pricing as Google Search.
Two things: metadata depth (siteName, creator, preview, favicon, image per source) and X integration — Grok cites X posts natively, so it's the only AI engine where you can measure how social posts versus web pages drive answers about your brand.
Yes — cited X posts appear as sources with the author in the creator field. Aggregating creators across your prompt set shows which voices shape Grok's view of your category.
Yes — searchQueries returns the queries Grok ran while researching. Combined with the sources array, you can trace the full retrieval path from prompt to citation.
The request shape is identical across engines — swap the endpoint and the same pipeline covers ChatGPT, Perplexity, Gemini, and Copilot.
The xAI developer API doesn't replicate the consumer product's live-search behavior or its citation surface. cloro drives the real grok.com product, so the sources you get are the sources users see.