Google Knowledge Panel Scraper

Overview

Extracts one entity record from the Google Search page for a query. The Actor reads the classic Knowledge Panel when Google renders it and also recognizes the current Overview module layout. It stores only fields visible in the selected Google entity panel, such as its name, summary, labeled facts, image, official website, source links, and related entities.

The exact fields vary by entity and locale. Missing facts are omitted. If Google does not render a usable panel or Overview module, the run fails instead of substituting data from another site or API.

The record includes counts for labeled facts, social profiles, related entities, and external sources. Source links include the visible label, publisher host, and URL; the official website is reported in website/officialWebsite and is not duplicated as a source link. A truncated fact value drops Google's trailing See more affordance. hasImage distinguishes a panel without an image from a missing image field. entityUrl is the panel's primary external entity reference (the summary source page when Google shows one, otherwise the official website or first panel source link), and retrieval.entityUrlSource records which of those Google actually rendered. retrieval.entityNameSource is search_query when an Overview module rendered no name of its own and the record uses the query as the entity name. Google redirect wrappers (/url?q=…, /goto?url=…, /interstitial?url=…) are resolved to the real destination, so no stored URL points at google.com.

Input

Field	Type	Default	Description
query	string	Albert Einstein	Entity to search for. Required.
gl	string	us	Google country code.
hl	string	en	Google interface language or locale tag, such as en-US.
maxRequestRetries	integer	10	Maximum retries for a temporary navigation failure (0–10).
requestTimeoutSecs	integer	90	Maximum request handler time in seconds (15–180), including the bounded wait for the hydrated panel.
navigationTimeoutSecs	integer	45	Maximum navigation response time in seconds (15–180). The Actor waits for the committed response, then for the panel content separately.
proxyConfiguration	object	Apify proxy	Optional browser-capable Apify proxy configuration. The HTTP-only GOOGLE_SERP group cannot tunnel a Chromium session.

The full input contract, including types, defaults, and constraints, is in .actor/input_schema.json.

Example:

{
  "query": "Albert Einstein",
  "gl": "us",
  "hl": "en"
}

Output

Each result is an entity record. attributes contains the visible label/value facts, while relatedEntities contains names from Google's related-entity area when that section is present. A Wikipedia source link is stored separately from social profiles. searchMetadata and retrieval identify the query, locale, page layout, and DOM-based extraction method.

Google renders the panel header before it hydrates the summary and the fact rows. The Actor therefore polls the panel subtree until it holds a summary, at least one labelled fact row, a website action, or an image (bounded by PANEL_POLL_ATTEMPTS), and it fails the request rather than storing a header-only record.

Optional fields are omitted when Google does not show them. The Actor does not manufacture a fixed field count: some entities have only a few facts, while others expose more than 20 useful values across their facts, related entities, and provenance.

The complete dataset field schema is in .actor/dataset_schema.json.

Example record

This example is based on the visible Google result for Albert Einstein. The fact text and related names below were visible in its Overview and “People also search for” sections; a live run may return a different set of optional fields.

{
  "recordType": "entity",
  "entityName": "Albert Einstein",
  "subtitle": "Theoretical physicist",
  "description": "Albert Einstein was a German-born theoretical physicist best known for developing the theory of relativity. Einstein also made important contributions to quantum theory. His mass–energy equivalence formula E = mc², which arises from special relativity, has been called \"the world's most famous equation\".",
  "entityUrl": "https://en.wikipedia.org/wiki/Albert_Einstein",
  "wikipediaUrl": "https://en.wikipedia.org/wiki/Albert_Einstein",
  "attributes": {
    "Born": "March 14, 1879, Ulm, Germany",
    "Died": "April 18, 1955 (age 76 years), Princeton Medical Center, Plainsboro Township, NJ",
    "Spouse": "Elsa Einstein (m. 1919–1936); Mileva Marić (m. 1903–1919)",
    "Children": "Eduard Einstein; Lieserl Einstein; Hans Albert Einstein",
    "Education": "University of Zurich (1905); ETH Zürich (1897–1900)",
    "Height": "5′ 9″",
    "Parents": "Hermann Einstein; Pauline Einstein"
  },
  "attributeCount": 7,
  "socialProfileCount": 0,
  "relatedEntityCount": 4,
  "sourceCount": 1,
  "hasImage": false,
  "sourceLinks": [
    {
      "label": "Wikipedia",
      "publisher": "en.wikipedia.org",
      "url": "https://en.wikipedia.org/wiki/Albert_Einstein"
    }
  ],
  "sourceUrls": ["https://en.wikipedia.org/wiki/Albert_Einstein"],
  "locale": { "country": "US", "language": "en" },
  "relatedEntities": [
    { "name": "Isaac Newton" },
    { "name": "Eduard Einstein" },
    { "name": "Elsa Einstein" },
    { "name": "Mileva Marić" }
  ],
  "searchQuery": "Albert Einstein",
  "type": "knowledge-panel",
  "resultType": "knowledge-graph",
  "searchUrl": "https://www.google.com/search?q=Albert%20Einstein&hl=en&gl=us",
  "searchMetadata": {
    "engine": "google-knowledge-panel",
    "query": "Albert Einstein",
    "country": "US",
    "language": "en",
    "scrapedAt": "2026-09-22T19:05:01.000Z"
  },
  "retrieval": {
    "method": "googleSearchDom",
    "source": "Google Search",
    "layout": "overview_module",
    "entityUrlSource": "description_source"
  },
  "scrapedAt": "2026-09-22T19:05:01.000Z"
}

Run and dataset

The Actor stores extracted entities in the default dataset. You can download the dataset in various formats such as JSON, HTML, CSV, or Excel.

The Actor uses a browser to render the public Google Search page. Navigation waits for the committed response (waitUntil: 'commit') because the Google SERP does not fire a reliable load event, and the knowledge panel is then awaited explicitly, so a slow third-party request cannot burn the whole navigation timeout. Browser and proxy usage affect run time. A page challenge, missing panel, or unavailable page ends the run without a fallback record.

Troubleshooting

• No panel record: Google intermittently answers the search URL with a JavaScript-bootstrap stub (tens of characters of body text, no result nodes) or with a result page whose panel module never rendered. Those pages are transient, so the query is retried with a fresh navigation until the panel appears or the retry budget is exhausted; the run only fails with EMPTY_PANEL after that. A page that did render a panel-like root but never hydrated is reported as PANEL_NOT_HYDRATED. Logs include bounded selector and content counts, plus the page title, without the search query or page text.
• Header-only panel: logs report reasonCode: PANEL_NOT_HYDRATED and the bounded polling attempt count; the record is not stored.
• Blocked page: The run fails with TARGET_BLOCKED when Google redirects to or renders a detected verification page. The terminal block is not retried; the Actor does not solve CAPTCHAs or switch to alternate data sources.
• Invalid input: The run fails with INVALID_INPUT when the query, locale, or bounded timeout/retry fields do not validate.
• Missing fields: Google varies the facts shown by entity, country, and language. The schema fields are optional where the source may omit them.

For support, include the Actor run URL and input used. Do not include credentials or private data in an issue.

Privacy and attribution

This Actor reads publicly displayed search content and stores the extracted entity facts in your dataset. Review the target site's terms and applicable laws before use. Google and any linked source sites are not affiliated with this Actor.