How much does a consensus search cost?

One page at each provider that returned results, added up: up to 3 credits. Providers that failed, came back empty, were blocked or never started are free, and a cache hit costs half. Your spend cap and key limit count the whole reservation.

Which providers take part?

Engines from the search route, one per index: for example Google or Brave, Bing (or a Bing-backed engine) and Wikipedia. meta.route lists every provider started and its outcome.

Is consensus available on news or images?

Not yet. engine: "consensus" works on /v1/search.

Are consensus results cached?

Yes, separately: a consensus answer is never served to a single-engine request, or the reverse.

engine=consensus

Consensus metasearch API

Ask up to 3 independent search indexes at once and get one merged list, ranked by how many of them agree. Each result says which engines returned it.

Get a free API key Consensus mode reference

How it works

Several indexes in parallel, merged and ranked by agreement.

engine: "consensus" on /v1/search works like a metasearch engine. Up to 3 providers answer in parallel, one per index: Yahoo and DuckDuckGo serve Bing's results, so at most one Bing-backed engine is used. Wikipedia's official API (the best matching article) can be one of the sources.

Results are deduplicated by URL (ignoring www., a trailing slash and tracking parameters such as utm_*) and ranked by how many providers returned them. Among equally agreed results, Wikipedia articles and results whose title and snippet contain your query's words come first, then the best position any provider gave. Each result carries sources, the providers that returned it.

A provider that fails or comes back empty hands its turn to the next index, so one flaky source doesn't leave you with a single-engine answer. It stops as soon as num unique results are in hand, or after a few seconds with whatever has arrived.

Consensus search

cURL

TypeScript

Python

curl https://api.serpkite.com/v1/search \
  -H "Authorization: Bearer $SERPKITE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"q":"best espresso machine","engine":"consensus"}'

Merged results · illustrative

{
  "results": [
    {
      "position": 1,
      "title": "The Best Espresso Machines",
      "link": "https://www.example.com/best-espresso-machines",
      "domain": "example.com",
      "sources": [
        "brave",
        "bing"
      ]
    },
    {
      "position": 2,
      "title": "Espresso machine - Wikipedia",
      "link": "https://en.wikipedia.org/wiki/Espresso_machine",
      "domain": "en.wikipedia.org",
      "sources": [
        "wikipedia"
      ]
    }
  ],
  "meta": {
    "engine": "consensus",
    "credits_used": 3,
    "route": [
      {
        "provider": "brave",
        "outcome": "ok",
        "ms": 640
      },
      {
        "provider": "wikipedia",
        "outcome": "ok",
        "ms": 410
      },
      {
        "provider": "bing",
        "outcome": "ok",
        "ms": 1180
      }
    ]
  }
}

Why it matters

What you get

Agreement as a signal

A page several independent indexes rank is a safer source for an agent than one engine's top hit.

Independent indexes

One provider per index, so two Bing-backed engines never count as two votes.

Clean dedupe

URLs are folded across www., trailing slashes and tracking parameters before ranking.

Resilient by design

A failed or empty provider hands its slot to the next index in the route.

Pay for answers only

Billing is one page per provider that returned results. Failed, empty and never-started providers are free.

Domain filters work

include_domains and exclude_domains go to every provider and apply to the merged rows.

FAQ