Verticals

POST /search/{vertical}

Beyond web search, Crisps returns structured data from specific sources — products, jobs, places, news, images, real-estate listings and YouTube transcripts — each normalized to a clean, stable schema so your agent doesn’t care which source produced it. Same auth, same credit metering as /search.

Verticals

Vertical	Path	Input	Returns
Products	`/search/products`	`query`	shopping items: title, price, rating, seller, url
Jobs	`/search/jobs`	`query`, `location`	postings: title, company, location, salary, apply links
Places	`/search/places`	`query`, `location`	businesses: address, phone, website, rating, hours, gps
News	`/search/news`	`query`	articles: title, url, source, date
Images	`/search/images`	`query`	images: title, image url, source
Real estate	`/search/realestate`	`url`	Zillow listings: address, price, beds/baths, estimates
YouTube transcript	`/search/youtube_transcript`	`url`	full transcript text + timed segments
Amazon	`/search/amazon`	`query`, `country`	marketplace items: title, price, rating, reviews, ASIN, url
Scholar	`/search/scholar`	`query`	papers: title, authors, snippet, citation count, url
Patents	`/search/patents`	`query`	patents: id, inventor, assignee, filing/grant dates, url
Trends	`/search/trends`	`query`, `country`	search interest over time: dated value series
Finance	`/search/finance`	`query`	quote for a ticker (`NVDA:NASDAQ`): price, movement, exchange
Bing	`/search/bing`	`query`	web results from Bing: title, url, snippet
DuckDuckGo	`/search/duckduckgo`	`query`	web results from DuckDuckGo: title, url, snippet
Brave	`/search/brave`	`query`	web results from Brave’s independent index

List verticals (products, jobs, places, news) are reranked for relevance to your query, best result first, with a score on each item.

Request

curl -X POST https://api.crisps.ai/search/products \
  -H "Authorization: Bearer $CRISPS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "query": "noise cancelling headphones", "max_results": 5 }'

import os, httpx
r = httpx.post(
    "https://api.crisps.ai/search/products",
    headers={"Authorization": f"Bearer {os.environ['CRISPS_API_KEY']}"},
    json={"query": "noise cancelling headphones", "max_results": 5},
)
print(r.json()["results"])

const res = await fetch("https://api.crisps.ai/search/products", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.CRISPS_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({ query: "noise cancelling headphones", max_results: 5 }),
});
const { results } = await res.json();

Try it

POST /search/products Try it

API base URL API key

Request body

Body parameters

Field	Type	Notes
`query`	string	Search term. Required for every vertical except `realestate` and `youtube_transcript`. For `finance` pass a ticker such as `NVDA:NASDAQ`.
`url`	string	Required for `realestate` (a Zillow search URL) and `youtube_transcript` (a video URL or id).
`location`	string	Optional locality hint (folded into the query) for `places`/`jobs`.
`country`	string	Two-letter country code. Selects the Amazon marketplace for `amazon` and the geo for `trends`.
`language`	string	Result language code.
`max_results`	integer	Cap on returned items (default 10). Ignored by `finance` (a single quote) and used as a rough cap by `trends`.

Response

Every vertical returns a uniform envelope: { "type": "...", "query": "...", "results": [ … ] }.

{
  "type": "products",
  "results": [
    { "title": "Sony WH-1000XM6", "url": "https://…", "price": "$399.99",
      "price_value": 399.99, "source": "Sony", "rating": 4.7, "reviews": 1203,
      "thumbnail": "https://…", "score": 0.83 }
  ]
}

{
  "type": "jobs",
  "results": [
    { "title": "Senior Python Engineer", "company": "Acme", "location": "Berlin, DE",
      "via": "LinkedIn", "highlights": {"schedule_type": "Full-time"},
      "apply": [{"title": "Apply on LinkedIn", "link": "https://…"}], "score": 0.79 }
  ]
}

{
  "type": "youtube_transcript",
  "video_id": "dQw4w9WgXcQ",
  "text": "We're no strangers to love …",
  "segments": [{ "text": "We're no strangers to love", "start": 0.15, "duration": 3.2 }]
}