Skip to content
ChangelogNew

Introducing Search

Search
A search for how heat pumps work in cold climates on POST /v1/search: three pages with their site logos and scores, the passage of the first page with the part that answers highlighted, and answer accuracy on FRAMES of 68.5% for Pro Search and 66.5% for Search

Give your agent web search in one call: each result comes with the passages of its page that answer the query, for 0.5 credits a search.

What you get

  • results: up to 10 pages for the query in score order, each with url, title, snippet and a relevance score from 0 to 1.
  • text: the passages of each page that answer the query, up to 10,000 characters a result, ready to pass to your model.
  • publishedDate and siteName when the page declares them.
  • The web_search tool on the querying.ai MCP server, at the same price, for Claude Code, Codex, Cursor and other MCP clients.

How it works

Send a query to POST /v1/search and the request stays open until the results arrive in the same response. Your agent gets its context from one request and skips the second call to fetch each page.

type picks the search. standard is the default. pro covers more of the web and picks passages more carefully for questions that need several sources, and answers in about four seconds. Query operators narrow either one: site:, -site:, after: and before:. A search still running after 30 seconds returns its task, and its results are read from GET /v1/async/task/{id}.

On 200 FRAMES questions on November 15, 2026, a model answering from Search results alone was correct 66.5% of the time, and 68.5% with Pro Search. The method and the full comparison are on the Search page: https://querying.ai/agent-search

Answer accuracy on FRAMES against price per 1,000 searches: querying.ai Search 66.5% at $0.50, Pro Search 68.5% at $2.00, Parallel fast 63.5% at $1.00, Parallel advanced 70.0% at $5.00, Exa 68.0% at $7.00, Tavily basic 48.5% at $5.00, Tavily advanced 53.5% at $10.00, Firecrawl 43.5% at $1.20
Cost of 1,000 correct answers on 100 company-news questions: querying.ai Search $0.51, Parallel fast $1.00, Firecrawl $1.26, querying.ai Pro Search $2.02, Parallel advanced $5.00, Tavily basic $5.10, Exa $7.07, Tavily advanced $10.20

Where it helps

  • Give a coding or research agent web search with the page text in one tool call.
  • Ground a chatbot’s answer in current pages and show the url of each passage as its source.
  • Keep a research run to trusted sites with includeDomains, or to recent pages with after:.

Send a request

Send the request with your API key in the Authorization header.

curl -X POST https://api.querying.ai/v1/search \
  -H "Authorization: Bearer $QUERYING_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "how do heat pumps work in cold climates",
    "type": "pro",
    "numResults": 3,
    "contents": { "text": { "maxCharacters": 500 } }
  }'

Request fields

querystringRequired
The search, up to 400 characters. Takes the operators site:example.com, -site:example.com, after:YYYY-MM-DD and before:YYYY-MM-DD.
typestring
standard or pro. pro searches deeper for questions that need several sources. Default standard.
numResultsinteger
How many results to return, 1 to 10. Default 5.
contents.text.maxCharactersinteger
Characters of page text each result carries in text, 200 to 10,000. Default 2,000.
userLocationstring
Two-letter country code, for results from one market. Leave it out for the fastest search, in the query’s language.
includeDomainsarray
Up to 20 host names. Results come only from these sites and their subdomains.
excludeDomainsarray
Up to 20 host names whose results are left out.
idempotencyKeystring
Your own ID for the request. Each key creates one task per account, so a retried send stays a single task.

Response

The response, trimmed for length.

JSON
{
  "success": true,
  "task": {
    "id": "…",
    "taskType": "SEARCH",
    "status": "COMPLETED",
    "createdAt": "…",
    "latencyMs": 4063
  },
  "credits": { "creditsToCharge": 2, "creditsCharged": 2 },
  "response": {
    "query": "how do heat pumps work in cold climates",
    "source": "all",
    "type": "pro",
    "results": [
      {
        "url": "https://www.carrier.com/us/en/residential/hvac-resources/heat-pumps/cold-climate-heat-pump/",
        "title": "Cold Climate Heat Pump Systems | Efficient Heating in Extreme …",
        "snippet": "In the winter, heat pumps work in cold weather by extracting thermal energy from the outdoor air and transferring it …",
        "text": "## Do cold climate heat pumps work?\n\nA heat pump is a versatile, year-round system that heats and cools your home by moving heat energy rather than generating it. In the winter, heat pumps work in cold weather by extracting thermal energy from the outdoor air and transferring it indoors.\n\n…",
        "score": 1,
        "publishedDate": "2026-09-15T11:03:00.000Z"
      },
      {
        "url": "https://www.mitsubishicomfort.com/articles/mythbusters-heat-pumps-for-extreme-cold-climates",
        "title": "Heat Pumps in Extreme Cold Climates | Mitsubishi Electric HVAC US",
        "snippet": "In cold climates, heat pumps typically experience a decrease in pressure, refrigerant flow rate and operational …",
        "text": "## Myth: Heat Pumps Are Only for Mild and Warm Climates\n\nHow do they work? In cold climates, heat pumps typically experience a decrease in pressure, refrigerant flow rate and operational capacity. …",
        "score": 0.8333,
        "publishedDate": "2026-10-19T12:06:06.108Z"
      },
      {
        "url": "https://homes.rewiringamerica.org/articles/heating-and-cooling/heat-pumps-cold-weather",
        "title": "Do heat pumps work in cold climates? - Rewiring America",
        "snippet": "…",
        "text": "In heating mode, heat pumps go into reverse and pump heat from the air outside your home to the inside. …",
        "score": 0.6667
      }
    ]
  }
}
task.statusstring
QUEUED, PROCESSING, then COMPLETED with the answer or FAILED with error.
credits.creditsChargedinteger
Credits charged for the finished task. A failed task charges 0.
response.resultsarray
Pages in score order, each with url, title, snippet, text and score, plus publishedDate and siteName when the page declares them.
response.results[].textstring
The passages of the page that answer the query, in page order, with headings as # lines and list items as - lines.
response.results[].scorenumber
Relevance from 0 to 1, comparable within one response.

Pricing

Credits are charged when a search completes: 0.5 for a standard search and 2 for a pro search. On the Max plan, 1,000 standard searches cost $0.50. A new account starts with 500 free credits, enough for 1,000 standard searches. Plans are listed at https://querying.ai/pricing .

More updates

Stay in the loop

Get new product updates in your inbox.

RSS feed