Google News Scraper: search, topics and local news as JSON avatar

Google News Scraper: search, topics and local news as JSON

Pricing

from $0.70 / 1,000 article scrapeds

Go to Apify Store
Google News Scraper: search, topics and local news as JSON

Google News Scraper: search, topics and local news as JSON

Google News as JSON from search queries (with operators like when:7d and site:), topic feeds and local feeds: headline, publisher, resolved publisher URL, publish time, source domain and related stories. No browser, no API key. From $1.00 per 1,000 articles.

Pricing

from $0.70 / 1,000 article scrapeds

Rating

0.0

(0)

Developer

subimpact

subimpact

Maintained by Community

Actor stats

0

Bookmarked

1

Total users

1

Monthly active users

6 hours ago

Last modified

Features

Limited permissions

Share

Google News RSS Scraper: Google News as JSON

Turn Google News into clean JSON rows. Give it search queries, topic headlines or place feeds, and get one row per article with the headline, publisher, the real publisher URL (not just a news.google.com link), publication time in ISO 8601 UTC, source domain and the cluster of related articles Google groups around the story.

No API key, no browser, plain HTTP. Google's own RSS feeds are the source, so every row is exactly what Google News shows — including the edition, language and ranking you ask for.

Why this one

  • Publisher URLs, resolved. Google News links are opaque ids (news.google.com/rss/articles/CBMi…). Every row also carries publisherUrl, the article on the publisher's own site, resolved through Google's public article endpoint — measured 100% on live runs.
  • Google search operators work. Quotes for exact phrases, site: to restrict a publisher, when:7d / when:1h for recency, -term to exclude, OR to combine. They are passed through to Google as-is.
  • Three feed types. Keyword search, the eight topic pages (World, Nation, Business, Technology, Entertainment, Sports, Science, Health) and geo feeds for places.
  • Cluster data. Topic and geo rows include relatedArticles: the other outlets covering the same story, with their own Google News links — useful for deduplication, source diversity and "who else reported this".
  • Any Google News edition. language + country select the edition (en-US/US, en-MY/MY, en-SG/SG, ms-MY/MY, id/ID, …), which changes both the articles and their language.

Typical uses: brand and PR monitoring, competitor tracking, a news dashboard or alerting pipeline, filling a newsletter or RAG store with linkable articles, and regional coverage where the local edition matters.

Input

{
"queries": ["\"electric vehicles\" when:7d", "site:reuters.com palm oil"],
"topics": ["BUSINESS", "TECHNOLOGY"],
"geos": ["Malaysia"],
"language": "en-MY",
"country": "MY",
"maxArticlesPerFeed": 50,
"maxItems": 200,
"resolveUrls": true
}
InputDefaultWhat it does
queriesprefill ["malaysia economy"]Google News search queries, one feed each. Search operators allowed.
topics[]Any of WORLD, NATION, BUSINESS, TECHNOLOGY, ENTERTAINMENT, SPORTS, SCIENCE, HEALTH.
geos[]Place names for geo feeds (Malaysia, Penang, Singapore, London).
languageen-USGoogle interface locale (hl): en-US, en-GB, en-MY, ms-MY, id, pt-BR, zh-CN, …
countryUSTwo-letter country for the edition (gl). With the language it forms Google's ceid (MY + en-MY → MY:en).
maxArticlesPerFeed100Rows taken per feed, in Google's ranking order. Google returns at most ~100 items per feed.
maxItems300Hard cap on delivered rows for the whole run (max 10,000).
resolveUrlstrueResolve publisherUrl for each article. Turn it off for a metadata-only run (faster, same price).
proxyConfigurationnoneOptional Apify Proxy, only needed if Google starts rate-limiting your runs.

Using only part of the input is fine — every combination of queries, topics and geos works, and a run with nothing selected stops with a clear error instead of scraping or charging anything.

What one row looks like

FieldMeaning
feedTypesearch, topic or geo — which Google News feed produced the row
feedThe query, topic or place you asked for (the value from queries / topics / geos)
feedUrlThe exact RSS URL fetched, incl. hl, gl, ceid — every row is reproducible
position1-based position inside the feed, i.e. Google's ranking
titleHeadline exactly as Google returns it (Google appends - <publisher>)
publisherPublisher name from Google's <source> element
publisherUrlThe article on the publisher's site, resolved from Google's id (null if it could not be resolved — googleNewsUrl is always there)
googleNewsUrlGoogle News link for the article
publishedAtPublication time, ISO 8601 UTC
snippetSummary text, HTML stripped. Google's standard feeds ship headline + publisher only, so this is usually null
sourceDomainPublisher domain from Google's <source url> (reuters.com, no www)
relatedArticlesCluster of other articles on the same story: [{title, publisher, url}], excluding this row's article. [] when Google clusters nothing
language, countryThe hl / gl the feed was requested with
scrapedAtWhen the row was built, ISO 8601 UTC
statusSuccess for a delivered article, or Empty / Error for a feed-level problem row (never charged)
urlResolvedByHow publisherUrl was obtained: batchexecute (Google's RPC), articleId (decoded from the id, no request), none, page (article page unusable), error, skipped (resolution time budget used), disabled (resolveUrls: false)
errorWhy an Empty / Error row has no article; null on Success rows

Example output

A real row from a live run ({"queries": ["malaysia economy"]}, 15 November 2026):

{
"feedType": "search",
"feed": "malaysia economy",
"feedUrl": "https://news.google.com/rss/search?q=malaysia%20economy&hl=en-US&gl=US&ceid=US%3Aen",
"position": 1,
"title": "Malaysia - Trade, Manufacturing, Services - Britannica",
"publisher": "Britannica",
"publisherUrl": "https://www.britannica.com/place/Malaysia/Economy",
"googleNewsUrl": "https://news.google.com/rss/articles/CBMiXkFVX3lxTE1sTVdHbUlkX0MzMndNMHg0SUI4OFdrQUlHVzM4NWxGblJ0QUl4aWFqNXYzSG0zaXlZSVFUX0hWb0hDeE9GZFh0LXNZT0FINHk5bmRCcXlyLVpieEJ3YXc?oc=5",
"publishedAt": "2026-11-12T08:00:00.000Z",
"snippet": null,
"sourceDomain": "britannica.com",
"relatedArticles": [],
"language": "en-US",
"country": "US",
"scrapedAt": "2026-11-15T08:45:45.510Z",
"status": "Success",
"urlResolvedBy": "batchexecute",
"error": null
}

A topic row from the same day ({"topics": ["BUSINESS"], "maxArticlesPerFeed": 4}), shortened to two related articles — this is the cluster Google puts around a story, minus the row's own article:

{
"feedType": "topic",
"feed": "BUSINESS",
"feedUrl": "https://news.google.com/rss/headlines/section/topic/BUSINESS?hl=en-US&gl=US&ceid=US%3Aen",
"position": 1,
"title": "FAA was warned for years about repair risks leading to nationwide ground stop - The Washington Post",
"publisher": "The Washington Post",
"publisherUrl": "https://www.washingtonpost.com/transportation/2026/11/14/faa-was-warned-years-about-repair-risks-leading-nationwide-ground-stop/",
"googleNewsUrl": "https://news.google.com/rss/articles/CBMixwFBVV95cUxQdGhEX0dMOEdKd2dQcXJYV1NJVnZKeUFMTlMyRmVMWXFkZHhaU3hDR0NCM201VTZQRXc0QXNERjdBdGdyX3ZXU3ZlbU42WjR4MW5qLWh1WGtCdk9UVXdlaUlQMTJkS1FuWHhDcUt0MlNmQlRRYTdDSTh0Z3I0RTAwaGlYdzByVG05WE1uZjdyeWhiQlYtRmRRSkwtOVR0WS14X1VOZ25JZzVHMFFJNEhMUW9YWWtKVTM3cG5McHJDbjJGWWFfMTZV?oc=5",
"publishedAt": "2026-11-15T04:37:00.000Z",
"snippet": null,
"sourceDomain": "washingtonpost.com",
"relatedArticles": [
{
"title": "US aviation shutdown blamed on L3 Harris installed router",
"publisher": "Reuters",
"url": "https://news.google.com/rss/articles/CBMiuAFBVV95cUxQWEQ1ckJVbXlQSHI5ZHU4VEppWVVZbXBYa0ZRWjdDdFNGWWxSdHF5bEtwR1Q2OHRxN2plYXctMm5lTU5TV0hzb2xQazg4U09MNDhyRnpOalhGYXJWN01mak9NR0dXWG0yR0k5WUtycmNhdkdZZkpYcXp5N2w0N2RQSUxEX1F5aEY1TGhDTzV5ZXl3dWR6QmdabDduOFBBWTNPdXBvVVY1VlAwVW1XYnIyQ2FwU2ZOeW5x?oc=5"
},
{
"title": "Sacramento airport travelers continue feeling effects of nationwide ground stop",
"publisher": "CBS News",
"url": "https://news.google.com/rss/articles/CBMilAFBVV95cUxPSHU5cnRUSGdWZloyUlU3Z24teWxuanR3QS13cUcwb290eDR4cHFjcm9majg2WC0wSzlKSU1SbU1CMFoxRWE0M0ozR2tUVENMamR0eUwxOFlRVDNJcFpIMTlXQS0yMnhDbGt2dGQ4RXE1eFV3TEJpOEhWaFRKYTZwSUgwcFZoTDcyLWttenJlU0NCcWpl?oc=5"
}
],
"language": "en-US",
"country": "US",
"scrapedAt": "2026-11-15T08:45:59.921Z",
"status": "Success",
"urlResolvedBy": "batchexecute",
"error": null
}

Search operators and editions

queries is passed to Google News, so anything Google News search accepts works:

  • "electric vehicles" — exact phrase.
  • site:reuters.com palm oil — one publisher only.
  • when:1h, when:7d, when:30d — recency.
  • palm oil -Indonesia — exclude a term.
  • (Malaysia OR Singapore) budget — combine.

language and country decide the edition. en-US + US is the international English edition; en-MY + MY is Malaysia's; ms-MY + MY is the Malay edition; id + ID is Indonesia's. Topic feeds follow the same edition, so NATION means "your country" for that edition.

Geo feeds resolve places Google knows for that country — Malaysia and Penang work with gl=MY, while an unknown or wrong-country place returns no items and one Empty row that is not charged.

Pricing

$1.00 per 1,000 articles — one article-scraped event per delivered row ($0.001), charged only after the row has been pushed. Empty and Error rows are free, and maxItems (default 300) keeps a run bounded: a default run of one query costs about 10 cents.

The comparison that matters: comparable Google News scrapers on the Store charge about $2–4 per 1,000 articles, mostly on top of their own platform usage. This actor is $1.00 per 1,000 with no start fee beyond Apify's $0.00005 minimum actor start event, and Apify platform usage (compute, and proxy if you switch it on) is paid by you on top — a 300-row run is normally a couple of minutes on 512 MB.

Limits and good to know

  • Google's RSS returns at most ~100 items per feed, so maxArticlesPerFeed cannot go above 100. For deeper history, schedule the actor (new items appear at the top) or use a news archive actor.
  • maxItems counts delivered rows and stops the run; feeds already fetched may add nothing after the cap.
  • Rows are not de-duplicated across feeds: the same article can appear in two queries, two topics and a geo feed. That is by design (each row records which feed found it) — dedupe on googleNewsUrl or publisherUrl if you merge feeds.
  • publisherUrl resolution sends one extra Google request per article and is best-effort: when Google does not answer (or after 10 minutes of resolving) the row is still delivered with googleNewsUrl and urlResolvedBy telling you why. resolveUrls: false skips it entirely.
  • Google's RSS descriptions normally contain no summary text, so snippet is usually null. Headlines keep Google's - <publisher> suffix in title; the clean name is in the publisher field.
  • Feed requests retry three times on 429/5xx with backoff. If Google starts throttling your scheduled runs, switch on proxyConfiguration (Apify Proxy) — plain HTTP measured clean on 15 November 2026: 25 article resolutions plus the feeds, no 429 and no 5xx.
  • Google publishes these feeds for feed readers with no key, but its copyright notice restricts the feed "for the purpose of rendering Google News results within a personal feed reader". You are responsible for how you use the data and for Google's terms.

For brand, keyword and competitor monitoring over a curated index of 350+ publisher news feeds — including Southeast Asian outlets that Google News surfaces late or not at all — use News Mentions API (subimpact/news-mentions-api). It returns publisher-native article rows with region, topic and language, and it is the better source when you need publisher feeds directly rather than Google's ranking. This actor is the better source when you want Google News search operators, Google's topic pages or an exact Google News edition.

Run monitoring and privacy

If a run fails or times out, this Actor tells its developer the run ID and status, so we can fix problems fast. It never sends your input, your results or your account data.