Google News Scraper

Pull headlines out of Google News the way you would read it: by keyword search, by topic section (Top stories, World, Business, Technology, Sports, …), by city / region (local news) or from any Google News URL you copy from the browser - in any of Google News' languages and country editions.

Every record is one unique story with the headline, publisher name and site, publish time, a short snippet, the other publishers covering the same story, and - unlike the Google News feed itself - the resolved publisher URL, so you can open or crawl the real article. URL resolution is included in the per-article price; there is no extra event for it.

Typical uses: media monitoring and brand alerts, competitor and industry news feeds, RSS replacement for dashboards and AI agents, local-news aggregation, trend and sentiment pipelines, feeding an article crawler with fresh publisher links.

What you get

One dataset record per story:

{
  "id": "CBMirAFBVV95cUxNYjF6M2x3R2hodlk0dGJGZGFKNmNZQlBEOVd1SXA2b2pGQk9wQnl2TnhZeldjZ25ZdzduTERsTTVfeDh5bzJpNXhKRTV6c1NwbnhPTjVjMGJBa1A0WVdWWUxfUHFPTVhLLXdoeVN0MWgtbVhGNGlsT2JzbVF3c3JQSk9XbXprNWIxZ3JhRlE5cW11ZU1US3Z4dmtYYXMwQ2ozY2lNM24tMGV4SXZZ",
  "title": "Exclusive: Longhorns used AI against Florida — a 'new wave' for UT sports",
  "url": "https://www.statesman.com/business/technology/article/texas-football-athletics-ai-bevo-lab-dell-22484462.php",
  "publisher": "Austin American-Statesman",
  "publisherUrl": "https://www.statesman.com",
  "publishedAt": "2026-10-23T14:16:41.000Z",
  "googleNewsUrl": "https://news.google.com/rss/articles/CBMirAFBVV95cUxN...?oc=5",
  "description": "Exclusive: Longhorns used AI against Florida — a 'new wave' for UT sports",
  "relatedArticles": [
    { "id": "CBMi...", "title": "Apple's Foldable iPhone Duo Arrives in Stores: Smartphone Review", "publisher": "The New York Times", "googleNewsUrl": "https://news.google.com/rss/articles/CBMi...?oc=5" }
  ],
  "urlResolution": "api",
  "source": { "type": "search", "value": "artificial intelligence", "feedUrl": "https://news.google.com/rss/search?q=artificial%20intelligence&hl=en-US&gl=US&ceid=US%3Aen" },
  "language": "en-US",
  "country": "US",
  "scrapedAt": "2026-10-24T19:39:05.910Z"
}

Field	Meaning
id	Google News story id (stable; used for deduplication)
title	Headline with Google's " - Publisher" suffix removed
url	Resolved publisher article URL, or null when resolution was off or failed
publisher, publisherUrl	Publisher name and homepage as shown by Google News
publishedAt	ISO 8601 UTC publish time
googleNewsUrl	The Google News redirect link
description	Up to 300 characters: the headline, or for clustered stories the related headlines and publishers
relatedArticles	Other publishers' coverage of the same story (empty when includeRelated is off or the story is not clustered)
urlResolution	api (resolved through Google News), offline (legacy id decoded locally), skipped (resolveUrls off) or failed
source	Which feed produced the record: search, topic, location or url, with the value and the feed URL
language, country	The Google News edition used

The key-value store also holds a SUMMARY record with feed counts, items seen, duplicates skipped, URL-resolution counts and whether maxItems was reached.

Input

Field	Type	Default	Notes
queries	string[]	["artificial intelligence"]	One Google News search per entry. Operators work: "exact phrase", OR, -word, site:reuters.com, intitle:, allintext:, when:7d, before:2026-01-01, after:2026-01-01. Each search feed returns up to ~100 newest stories.
topics	enum[]	[]	Headline sections: TOP, WORLD, NATION, BUSINESS, TECHNOLOGY, ENTERTAINMENT, SPORTS, SCIENCE, HEALTH.
locations	string[]	[]	Local-news feeds, e.g. Chicago, Sydney NSW, Bavaria.
startUrls	request list	[]	Google News search / topic / section / RSS URLs copied from the browser; converted to RSS automatically. Other URLs are ignored (listed in the log and SUMMARY).
language	string	en-US	Google hl: en-US, en-GB, de, fr, es, pt-BR, ja, hi, …
country	string	US	Google gl; empty derives it from the language. Together they pick the edition (ceid).
timeRange	enum	any	1h, 1d, 7d, 1m, 1y - applied to search queries only, unless the query already has when:.
maxItems	integer	100	Total unique stories per run (exact).
maxItemsPerFeed	integer	100	Cap per query / topic / location / URL. Google returns at most about 100 per feed.
resolveUrls	boolean	true	Resolve each Google News link to the publisher URL (two lightweight requests per story, no extra charge). Off = faster, Google links only.
includeRelated	boolean	true	Keep relatedArticles.
proxyConfiguration	object	Apify proxy	Datacenter proxies are fine for normal volumes; use residential if Google rate-limits very large runs.

At least one of queries, topics, locations or startUrls is required.

Example inputs

Media monitoring, last 24 hours, UK edition:

{ "queries": ["\"Acme Corp\" OR \"Acme Corporation\"", "Acme recall"], "language": "en-GB", "country": "GB", "timeRange": "1d", "maxItems": 200 }

German business and tech headlines plus Bundesliga, Google links only (fastest):

{ "topics": ["BUSINESS", "TECHNOLOGY"], "queries": ["Bundesliga"], "language": "de", "country": "DE", "resolveUrls": false }

Local news for several cities:

{ "locations": ["Chicago", "Austin TX", "Denver"], "maxItemsPerFeed": 30, "maxItems": 90 }

Pricing

Pay per event. $0.005 per run start plus $0.002 per story returned ($2 per 1,000), with publisher URL resolution and related coverage included. A default run of 100 stories costs about $0.21. Compared with other Google News Actors on the Store (typically $0.002–0.005 per item, some charging extra for resolved URLs), this is at the low end with the resolved URL built in.

How it works and limitations

• The Actor reads Google News' public RSS feeds (search, section, local and topic feeds) and, when resolveUrls is on, resolves each story link through Google News' own redirect service to the publisher URL. No login, cookies or CAPTCHA solving are involved.
• Google News caps feeds at roughly 100 stories each. For more coverage, split a broad query into several narrower ones or use when: / before: / after: ranges, and use locations and topics alongside queries.
• maxItems is exact and filled in the order feeds finish; use maxItemsPerFeed to balance several feeds.
• Output is facts and metadata only: headline, publisher, time, links and a snippet of at most 300 characters. Article bodies and images are not copied; use the url with an article crawler if you need full text.
• Stories are deduplicated by Google's story id within a run (the same story appearing in two feeds is returned once, attributed to the first feed that produced it).
• Google occasionally rate-limits an IP; the Actor retries with a new session and proxy. If a feed still fails after 3 retries it is counted in SUMMARY.feedsFailed, and the run fails only when every feed failed.

Get alerted when records change

Pair this Actor with Dataset Diff & Alert Monitor to receive only the new, changed or removed records in Slack, Discord, Telegram or a webhook. Schedule the monitor, not this Actor: it runs this Actor for you, compares runs by id, and a run that finds nothing costs only the comparison. Scheduled hourly this is a Google Alerts replacement you control: your keywords, your edition, your channel, at $0.002 per new story plus the comparison.

{
  "sourceActorId": "mdcurt/google-news-scraper",
  "sourceInput": {
    "queries": [
      "\"Acme Corp\" OR \"Acme Corporation\""
    ],
    "timeRange": "1d",
    "maxItems": 100
  },
  "keyFields": [
    "id"
  ],
  "notifyOn": [
    "added"
  ],
  "labelField": "title",
  "urlField": "url",
  "slackWebhookUrl": "https://hooks.slack.com/services/…"
}

Changelog

• 1.0.0 (2026-10-24) - Initial release: search, topic, location and start-URL feeds; language/country editions; time range; publisher URL resolution (API + offline legacy ids); related coverage; exact maxItems; SUMMARY record.