Can I filter by size, colour or licence?

Not yet. The endpoint takes a text query, country, language and page.

Does DuckDuckGo have a search API?

DuckDuckGo's own Instant Answer API returns topic summaries, not the web results list. SerpKite reads DuckDuckGo's public results and returns the full list as JSON. It is not affiliated with DuckDuckGo.

Are DuckDuckGo web results from Bing?

Largely, yes. DuckDuckGo's organic web results come mostly from Bing's index, which is why engine=consensus counts DuckDuckGo and Bing as one index. News, images and videos come from DuckDuckGo's own feeds.

How much does a DuckDuckGo search cost?

1 credit per page, the same as Google, from $0.60 per 1,000. Failed, blocked and empty requests are free.

POST /v1/images · engine=duckduckgo

DuckDuckGo Image Search API

DuckDuckGo's image results with full-size URLs, thumbnails, dimensions and the page each image comes from. One credit usually returns 50 to 90 images.

Updated Oct 25, 2026

Get a free API key Try in playground

1 credit per page of results 2,500 free credits, no card Failed calls are free

Overview

What the DuckDuckGo Image Search API does

DuckDuckGo's web results come largely from Bing's index; its news, images and videos come from its own feeds. Use it for DuckDuckGo's results without scraping it yourself, or as a fallback with broad image and video coverage.

/v1/images with engine: "duckduckgo" returns DuckDuckGo's image results: image_url, thumbnail_url, image_width, image_height, and the link, domain and source of the host page.

What you get

results[] array: Images: position, title, image_url (full size), thumbnail_url, image_width, image_height, and the host page's link, domain and source.
meta object: request_id, credits_used, cached, latency_ms, parse_quality, engine (the provider that answered) and route (each provider attempt with outcome and ms).

Request

POST /v1/images

cURL

Python

TypeScript

fetch

PHP

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

Response (live response, trimmed)

200 OK · DuckDuckGo images · live response, trimmed

{
  "request": {
    "country": "us",
    "endpoint": "images",
    "engine": "duckduckgo",
    "language": "en",
    "num": 10,
    "page": 1,
    "q": "espresso machine"
  },
  "results": [
    {
      "domain": "tiendamia.com",
      "image_height": 1415,
      "image_url": "https://m.media-amazon.com/images/I/81ABm7+MerL._AC_SL1500_.jpg",
      "image_width": 1500,
      "link": "https://tiendamia.com/ar/producto?amz=B0DJSLMTWY",
      "position": 1,
      "source": "tiendamia.com",
      "thumbnail_url": "https://tse2.mm.bing.net/th/id/OIP.nuxCLYiFGF3PicOuOkmiugHaG_?r=0&pid=Api",
      "title": "Espresso Machine with Built-in Grinder,Professional 15 Bar Espresso ..."
    },
    {
      "domain": "madtasting.com",
      "image_height": 575,
      "image_url": "https://madtasting.com/wp-content/uploads/2024/08/top_espresso_machines_2024-16.jpg",
      "image_width": 1006,
      "link": "https://madtasting.com/vetted/best-commercial-super-automatic-espresso-machine/",
      "position": 2,
      "source": "madtasting.com",
      "thumbnail_url": "https://tse1.mm.bing.net/th/id/OIP.ZGckgsDI_cvaMuwsSFNVTgHaEO?r=0&pid=Api",
      "title": "15 Best Commercial Super Automatic Espresso Machines of 2025 - Brew ..."
    }
  ],
  "meta": {
    "request_id": "req_01J8ZK4M6Q2V7",
    "credits_used": 1,
    "cached": false,
    "engine": "duckduckgo",
    "route": [
      {
        "provider": "duckduckgo",
        "outcome": "ok",
        "ms": 2455
      }
    ],
    "latency_ms": 2475,
    "parse_quality": "ok"
  }
}

Every response also carries X-Credits-Used, X-Credits-Remaining, X-Cache and X-Tokens-Estimate headers. Run this query in the playground

One key, many engines

DuckDuckGo through SerpKite

Same key, same envelope and the same SDK call as Google: only engine changes.
meta.engine names the provider that answered and meta.route lists every attempt. A result is never relabelled.
engine: "auto" keeps Google first and falls back to DuckDuckGo and the other engines only when Google fails. How fallback works
Failed, blocked and empty attempts are free; you pay only for the provider that answered.

Everything DuckDuckGo serves on SerpKite

DuckDuckGo Search API POST /v1/search · engine=duckduckgo
DuckDuckGo News API POST /v1/news · engine=duckduckgo
DuckDuckGo Image Search API POST /v1/images · engine=duckduckgo
DuckDuckGo Video Search API POST /v1/videos · engine=duckduckgo

See all engines on the multi-engine search page or in the providers reference.

engine=auto answered by DuckDuckGo (illustrative)

meta.route

{
  "request": {
    "endpoint": "images",
    "engine": "auto",
    "q": "espresso machine"
  },
  "meta": {
    "engine": "duckduckgo",
    "route": [
      {
        "provider": "google",
        "outcome": "blocked",
        "ms": 2140
      },
      {
        "provider": "duckduckgo",
        "outcome": "ok",
        "ms": 640
      }
    ]
  }
}

Measured Oct 25, 2026

How DuckDuckGo behaves on this endpoint

Three queries ("espresso machine", "northern lights norway", "mid century modern chair") sent to /v1/images with engine=duckduckgo through the production API.

Images per page: 57 to 92, depending on the query
Median response time: 2.8 s
Fields on every image: title, image_url, thumbnail_url, image_width, image_height, link, domain and source
Image size: Median width 1,200 px; 83% at least 1,000 px wide
Host sites: 117 different domains across 228 images

Filters and limits

Images come from DuckDuckGo's JSON image feed, which needs a per-query token. SerpKite gets the token for you; it's reused for later pages of the same query.
time works for day, week, month and year. time: "hour" and custom date ranges are a 400 invalid_request and aren't charged.
country and language pick one of DuckDuckGo's regions; a country without one searches worldwide.
On the same three queries, DuckDuckGo returned fewer images per page than Brave (228 against 300) but larger ones, from twice as many sites (117 against 56). Use it when you want big, varied images; use Brave when you want the most images per credit.
One page costs 1 credit however many images it holds.

Reference

Parameters

The parameters /v1/images accepts, in the JSON body or the query string.

DuckDuckGo Image Search API parameters
Name	Type	Default	Description
`q` required	string	–	The search query. Required. Up to 2,048 characters.
`engine`	string \| string[]	google	"duckduckgo" for this page. "auto" tries Google first and falls back to other engines; a list such as ["google", "duckduckgo"] limits the route to those engines.
`country`	string	us	Country to search from, as a two-letter ISO code (us, gb, de, in…). Defaults to the country a location ends in.
`language`	string	en	Interface language, as a language code with an optional region (en, de, pt-BR, zh-TW…).
`page`	integer	1	Results page, 1–10. Each page is billed separately.
`format`	string	json	Response format. markdown is LLM-ready prose; compact is JSON with only the fields agents need. Values: json, compact, markdown.
`fields`	string	–	Comma-separated projection, e.g. results.title,results.link,knowledge_graph. Cuts tokens.
`max_age`	integer	–	Accept a cached result up to this many seconds old. Cache hits cost 50% of the credits.

Full reference, error codes and headers are in the /v1/images reference. Queue up to 100 requests at half price with POST /v1/batches.

Use cases

What people build with the DuckDuckGo Image Search API

Media search

DuckDuckGo images and videos return large result sets per page, with dimensions, durations and channels.

Fallback for agents

engine=auto answers from DuckDuckGo when Google and the engines before it are unavailable.

Privacy-minded products

Show DuckDuckGo's results in a product that promises users a non-Google source.

Pricing