Web Search

Get web search results as JSON from Google or DuckDuckGo in one API call: organic results with title, URL, snippet, and position. Filter by date and domain. Set country and language preferences. No proxy or browser to manage.

Web Search answers HTTP requests directly. Use it as a search tool for an AI agent or LLM app, or call it from any script or backend.

Why use Web Search?

• Give an AI agent a search tool. Results come back as JSON the model can read, in the same request. Ground answers in current web content without a results page to parse.
• Control search scope and locale. Set date and domain filters, and choose a country and language for ranking.
• No infrastructure to run. Apify handles hosting, proxies, and IP rotation. Call the endpoint.
• Call it from anywhere. Use HTTP or MCP for direct requests. For regular Actor runs, use the Apify API client libraries for Python and JavaScript, or Apify integrations like Make, Zapier, and n8n.

How much does it cost?

Web Search uses pay-per-event pricing: one event per successful search, including searches with no results. maxResults doesn't change the price per search. Failed searches (invalid input, blocked, timed out) aren't charged a search event.

A regular Actor run also charges an Actor Start event. See the Pricing tab for current rates.

Getting started

1. Use your Apify API token in a token query parameter or an Authorization: Bearer <token> header. The header is safer: URLs can appear in logs and browser history.

2. Send a GET or POST request to https://web-search.apify.actor with your query. Both accept the same fields:

   $curl 'https://web-search.apify.actor/?query=latest+AI+agent+payment+protocols&maxResults=5&token=***'
   

   curl -X POST 'https://web-search.apify.actor/' \
     -H 'Content-Type: application/json' \
     -H 'Authorization: Bearer ***' \
     -d '{
       "query": "latest AI agent payment protocols",
       "maxResults": 5,
       "allowedDomains": ["wikipedia.org", "github.com"]
     }'
   

   Replace *** with your Apify API token. The POST example restricts results to Wikipedia and GitHub. Omit allowedDomains to search across domains.

3. Read results from the JSON response. No polling, no separate results endpoint.

For a regular Actor run, enter the same fields on the Input tab and select Start. Web Search searches once, saves the result to the run's dataset, and exits. Regular runs also work with Apify integrations and scheduled searches.

Input

Send input to https://web-search.apify.actor as GET query parameters or a POST JSON body. For a regular Actor run, enter it on the Input tab. query is the only required field.

Field	Description
query	Search text (required). Operators pass through unchecked.
provider	google (default) or duckduckgo. See Choosing Google or DuckDuckGo.
maxResults	Maximum results to return. Defaults to all results on one page. Raising it won't fetch more pages.
languageCode	ISO 639-1 code with an optional region suffix, like en or pt-BR. Default: en.
countryCode	ISO 3166-1 alpha-2 country code, like US, DE, or CZ. Default: US. Unsupported markets return INVALID_COUNTRY.
location	Google-only place name, most specific first, like Paris, Ile-de-France, France. Best-effort.
dateFrom / dateTo	Inclusive publication-date bounds in YYYY-MM-DD format. Best-effort.
allowedDomains / blockedDomains	Up to 10 hostnames per list. Comma-separated in GET, a JSON array in POST.

When setting both date bounds, dateFrom must not be later than dateTo. Date filters rely on the search engine's publication-date signals, which can be missing or wrong.

Choosing Google or DuckDuckGo

Use Google to target a location or get more result details. Try DuckDuckGo for a second ranking, or if Google finds nothing or fails. DuckDuckGo returns one page of about 10 results.

	Google	DuckDuckGo
location	yes	no
countryCode	192	57
dateFrom / dateTo	yes	yes
allowedDomains / blockedDomains	yes	yes
emphasizedKeywords	yes	yes
totalResults, correctedQuery, suggestedQuery	yes	no
date, siteName, siteLinks	yes	no

You can query both providers in parallel to widen coverage or cross-check results. Each search is charged as its own event.

Locale fields

Use an ISO 639-1 languageCode like en, de, or cs. A region suffix works too, as in pt-BR. Google uses languageCode to localize its interface and favor pages in that language. It doesn't translate results or strictly filter them by language.

countryCode runs the search as if you were in that country. This changes result order and which local businesses appear. Google uses the matching domain (google.de for DE). DuckDuckGo uses the matching region, preferring the languageCode variant (CA with fr) over the default. Results aren't filtered by hosting country.

location sets the searcher's presumed location for local and "near me" results. It doesn't filter the pages themselves. DuckDuckGo has no location parameter, so provider: "duckduckgo" rejects it with INVALID_PARAMETER.

Domain filters

A domain matches itself and all of its subdomains, so wikipedia.org also matches en.wikipedia.org. A bare top-level domain works too: gov matches every domain under it. When set, allowedDomains keeps matching domains and blockedDomains excludes them. Both default to empty lists. Filters can remove every result on the page. If results is empty, loosen the filters before changing the query.

Output

A successful search returns JSON with query, a search object with search details, and a results array of organic results. An empty results array still counts as a successful search. A regular Actor run saves this JSON as one dataset item. Download it as JSON, HTML, CSV, or Excel from the Output tab.

Example response for a search without domain filters:

{
    "query": "latest AI agent payment protocols",
    "search": {
        "provider": "google",
        "searchedAt": "2026-08-27T11:42:18.512Z",
        "durationSecs": 2.104,
        "correctedQuery": null,
        "suggestedQuery": null,
        "totalResults": 1240000,
        "countryCode": "US",
        "languageCode": "en",
        "location": null,
        "providerDomain": "google.com",
        "dateFrom": null,
        "dateTo": null,
        "checkUrl": "https://www.google.com/search?q=latest+AI+agent+payment+protocols&gl=us&hl=en"
    },
    "results": [
        {
            "position": 1,
            "url": "https://example.com/agent-payments",
            "domain": "example.com",
            "title": "Agent Payments Protocol",
            "snippet": "A protocol for authorizing payments by autonomous agents.",
            "siteLinks": [{ "title": "Specification", "url": "https://example.com/agent-payments/spec" }]
        }
    ]
}

If the search fails (invalid input, blocked, timed out), Web Search returns an error with code, error, and a matching HTTP status. See Troubleshooting for recovery steps:

{
    "code": "INVALID_DOMAINS",
    "error": "\"allowedDomains\" accepts at most 10 domains, got 12."
}

See the Actor's Endpoints tab for the full API reference.

Response

Field	Description
results[].position	1-based rank within results, counted after filtering. Always dense: 1, 2, 3, ...
results[].url	Absolute destination URL, never a search engine redirect.
results[].domain	Hostname of the destination URL, including subdomains.
results[].title	Result heading.
results[].snippet	Result excerpt. null when the search engine shows none.
results[].displayedUrl	Breadcrumb string, when available.
results[].siteName	Publisher name, when available.
results[].date	Publication date (YYYY-MM-DD), when available.
results[].emphasizedKeywords	Bolded terms inside the snippet, omitted when there are none.
results[].siteLinks	Sub-page links attached to the result, omitted when there are none.
search.provider	Search engine used.
search.providerDomain	The engine domain queried, like google.de for countryCode: "DE". Always duckduckgo.com for DuckDuckGo.
search.totalResults	The engine's estimate of all matching pages, not the number returned. null if not shown, and always for DuckDuckGo.
search.correctedQuery	The query the search engine silently searched instead. Always null for DuckDuckGo.
search.suggestedQuery	The engine's "did you mean" suggestion, not applied to the search. Always null for DuckDuckGo.
search.checkUrl	Link to the query as sent to the engine. Results may differ: engines personalize by IP, account, and time.

Search via MCP

Web Search runs a Model Context Protocol server at /mcp. Its only tool, search, accepts the same inputs as the REST API. It returns the search JSON in an MCP text content block and marks search failures with isError.

For Claude Code, add the HTTP endpoint with your Apify API token:

$claude mcp add web-search https://web-search.apify.actor/mcp -t http -H 'Authorization: Bearer ***'

Replace *** with your Apify API token. For another MCP client, like Cursor, use the same endpoint and an Authorization: Bearer <token> header.

Or open the Apify MCP server setup page with Web Search selected. Copy the settings for your client.

Troubleshooting

Common errors

For input validation errors, find the field named in error under Input.

Error code	When it happens	What to do
MISSING_QUERY, INVALID_*	A field is missing or invalid, the JSON body is malformed, or location was sent with provider: "duckduckgo".	Fix what error describes. For DuckDuckGo, remove location and use countryCode.
BLOCKED_BY_SEARCH_ENGINE	The search engine served a consent or CAPTCHA page, or refused the request.	Retry the request, or switch provider.
SEARCH_TIMEOUT	The search exceeded its 60-second budget.	Retry the request.
UPSTREAM_SEARCH_ERROR	The search backend couldn't be reached or returned an unexpected status.	Retry the request, or switch provider.
PARSE_ERROR	Web Search couldn't recognize the search results page, which can happen when its markup changes.	This is a bug on our side. Report it on the Issues tab.
INTERNAL_ERROR	The request failed unexpectedly.	Retry the request. If it keeps failing, report it on the Issues tab.

Need support or a custom feature?

If a retry doesn't help, open an issue on the Actor's Issues tab. Include your input and the error, or chat with Apify support. For different fields, another search engine, or a different result shape, ask Apify for a custom solution.

FAQ

Does Web Search return ads or featured snippets?

No. Only organic results are returned.

Are site: and filetype: operators supported?

Yes, but Web Search passes them to the search engine unchecked. Use allowedDomains and blockedDomains for enforced domain filtering.

Why do Web Search and browser results differ?

Search engines personalize by IP, account, and time. checkUrl shows the query as sent, not the same result set.

Is scraping search results legal?

Scraping publicly available search results is generally accepted, but you are responsible for complying with the search engine's terms of service and any laws that apply to your use case. If you're unsure, seek legal advice. Read more in this blog post.