Everything on the results page, as structured data

One REST API with four endpoints. Same authentication, same response envelope, same credit model.

GET /api/v1/search

Organic results with title, URL, displayed link, snippet, source and date. Up to 100 results per call, plus related searches.

GET /api/v1/images

Original image URL, thumbnail, source page and pixel dimensions, up to 150 per call.

GET /api/v1/autocomplete

Live query suggestions for any prefix, in the language and market you choose.

GET /api/v1/rank

Pass a keyword and a domain, get its position in the top 100 and every URL from that domain that ranks.

Beyond search results

The same key also calls 172 endpoints on 60 platforms: social and video, marketplaces, maps and travel, real estate, job boards, app stores and developer sites. One credit per request.

GoogleBingYouTubeTikTokAmazonWalmartYelpTripadvisorZillowRedfinIndeedGreenhouseAppleSteamGitHubStack Overflow
Browse the APIs

Targeting and pagination

Control market, language, depth and freshness on every request.

qSearch query, up to 400 characters. Quotes and operators such as site: work.
glCountry as a 2-letter code. Defaults to US.
hlInterface and result language. Defaults to the country's main language.
numResults per call, 10 to 100. Each 10 results costs one credit.
pagePage number for paginating through results.
timeFreshness filter: past day, week, month or year.

Response fields

Every organic result includes:

organic_results[0]
{
  "position": 1,
  "title": "The 15 Best Running Shoes of 2026, Tested by Runner’s World",
  "link": "https://www.runnersworld.com/gear/a19663621/best-running-shoes/",
  "displayed_link": "https://www.runnersworld.com › gear › best-running-shoes",
  "source": "runnersworld.com",
  "snippet": "Find the best running shoes of 2026, including Runner’s World-tested picks from Adidas, Brooks, New Balance, Nike, …",
  "date": "Aug 3, 2026"
}

Reliability built in

  • Search engines sometimes answer automated traffic with unrelated results instead of an error. serpfast checks every page for relevance and retries before you see it.
  • Requests that fail upstream are refunded automatically and logged with a clear error code.
  • Rate limits are per API key, so one busy integration cannot slow down another.

A dashboard that shows where credits go

Create and revoke keys, run queries in the playground and review every request with its status, latency and cost.

Create free account