Google News Scraper – Headlines, Real Article URLs & Alerts

Scrape Google News headlines for any keyword, topic, location, language and country, and get the real publisher URL for each one instead of a news.google.com redirect link. It reads Google News RSS feeds over plain HTTP (no browser, no login), so runs are fast and cheap.

• Any query, topic or edition. Google News operators work ("exact phrase", OR, -exclude, site:, when:), as do topics, local news, top stories and any language/country edition, with time windows from 1 hour to 30 days.
• Real article URLs and alerts. Google News links are decoded to the publisher URL, duplicates across queries are removed, and monitor mode returns only articles you haven't seen yet.
• $1 per 1,000 articles. Optional og:image/description metadata costs +$0.002 per article. The Actor returns headlines, links and metadata only, never full article text.

Quick start

This is the prefilled input: 5 headlines in a few seconds, for about $0.005.

{ "queries": ["artificial intelligence"], "maxItemsPerQuery": 5 }

Sample output

title	source_name	published_at	url_resolved
‘That’s so AI!’ What gen Alpha’s biggest insult tells us	The Guardian	2026-09-24	true
US, China agree to cut tariffs on $30 billion worth of goods, set up channel for AI incidents	The Hill	2026-09-26	true
Agentes no autorizados de OpenAI atacaron tres sitios web distintos del Gobierno de EE.UU.	CNN en Español	2026-09-26	true

Price comparison (Apify Store, September 2026)

Actor	Price per 1,000 articles	Monthly users
Google News Scraper (this Actor)	$1 (+ $0.0001 per run)	new
data_xplorer/google-news-scraper-fast	$4	557

Related Actors: Bluesky Scraper, Substack Scraper and Polymarket Scraper & API.

What can you use it for?

• Media monitoring & PR: track mentions of your brand, competitors or executives in every country
• News alerts: schedule monitor mode every hour and send only new headlines to Slack, email or a webhook
• Market & finance research: headlines about tickers, sectors or commodities in real time
• AI agents & RAG: give an LLM fresh, dated headlines with publisher links to cite
• Local news: collect regional headlines by city or state
• Datasets: build headline datasets across languages for NLP and trend analysis

Input example

{
  "queries": ["artificial intelligence", "\"OpenAI\" OR \"Anthropic\" -stock"],
  "topics": ["TECHNOLOGY"],
  "language": "en",
  "country": "US",
  "timeWindow": "1d",
  "maxItemsPerQuery": 50,
  "resolveOriginalUrl": true,
  "fetchArticleMetadata": false,
  "monitorMode": false
}

Field	Description
queries	Keywords (one per line). Google News search operators work.
topics	WORLD, NATION, BUSINESS, TECHNOLOGY, ENTERTAINMENT, SPORTS, SCIENCE, HEALTH, or a topic ID / news.google.com/topics/... URL.
locations	City/region names for local headlines, e.g. Berlin, Texas.
includeTopStories	Also scrape the edition's top stories.
language / country	Edition, e.g. en/US, en/GB, de/DE, tr/TR, es-419/MX, pt-419/BR, ja/JP.
timeWindow	any, 1h, 1d, 7d, 30d.
maxItemsPerQuery	1–100 (Google News RSS returns up to ~100 per feed).
dedupe	Remove duplicates across queries (article ID, headline+source, publisher URL). Default on.
resolveOriginalUrl	Decode to the publisher URL. Default on.
fetchArticleMetadata	Read og:* metadata from the publisher page (+$0.002 per enriched article).
monitorMode / monitorStateKey	Only new articles since previous runs; state lives in the named key-value store google-news-monitor.
proxyConfiguration	Optional Apify datacenter proxy (normally not needed).

Output example

{
  "title": "‘That’s so AI!’ What gen Alpha’s biggest insult tells us",
  "source_name": "The Guardian",
  "source_url": "https://www.theguardian.com",
  "published_at": "2026-09-24T04:00:00Z",
  "article_url": "https://www.theguardian.com/society/2026/sep/24/thats-so-ai-what-gen-alphas-biggest-insult-tells-us",
  "url_resolved": true,
  "google_news_url": "https://news.google.com/rss/articles/CBMioAFBVV95cUxQ...?oc=5",
  "snippet": null,
  "image_url": null,
  "query": "artificial intelligence",
  "feed_type": "search",
  "rank": 1,
  "language": "en",
  "country": "US",
  "article_id": "CBMioAFBVV95cUxQ...",
  "scraped_at": "2026-09-27T01:12:42Z",
  "related_coverage": null
}

With fetchArticleMetadata: true, items also carry meta_title, meta_description, meta_image, meta_published_time, meta_site_name, meta_author and metadata_fetched. snippet and image_url are then filled from og:description / og:image. Top stories include related_coverage, a list of other outlets covering the same story. A run summary (resolution stats, per-feed counts, errors) is saved to the OUTPUT record of the default key-value store.

Pricing

Pay per event, so you only pay for articles you get:

Event	Price
Article	$0.001 ($1 per 1,000)
Metadata enrichment (optional, only when found)	+$0.002
Actor start	$0.0001

Duplicates, already-seen articles in monitor mode, failed feeds and blocked publisher pages cost nothing. The Actor stops cleanly when your run's maximum charge is reached.

	This Actor	Market leader (Apify Store, Sept 2026)
Price per 1,000 articles	$1	$4
Original publisher URL	✅ decoded	varies
Monitor mode (only new)	✅ built-in	–
Dedupe across queries	✅	–

Use with AI agents / Apify MCP

The Actor works as a tool for Claude, ChatGPT, Cursor and other MCP clients through the Apify MCP server. Add it by name and let the agent call it with {"queries": ["<topic>"], "timeWindow": "1d", "maxItemsPerQuery": 20}. The output is flat JSON with ISO dates and publisher links the agent can cite. The input is small and every field has a default, so agents rarely send a bad request. For recurring briefings, schedule it with monitorMode: true so each run returns only new headlines.

FAQ

How many articles can I get per query? Google News RSS returns up to about 100 articles per feed. For more coverage, split a topic into narrower queries (e.g. add site: or when: / after: operators) or run several time windows. Results are deduplicated across queries.

How does original URL resolution work? Google News links are encoded. The Actor decodes them with the same public endpoints news.google.com uses in the browser. Resolution is best-effort. If Google changes or throttles it, you still get the article with the Google News URL in article_url and url_resolved: false. The run doesn't fail.

Do I need a proxy? Usually not. Tests resolved 600 articles in one run without a proxy. If you run very large jobs and see url_resolved: false, enable the Apify datacenter proxy or lower maxConcurrency.

Does it return the full article text? No. It returns headlines, links, dates and page metadata (og:title/description/image) only. It never returns full article text.

Why is some metadata missing? Some publishers block automated requests (HTTP 403) or have no og tags. The reason is shown in metadata_error, and you're only charged for metadata when it's found.

How does monitor mode remember what I've seen? It stores article IDs, headline+source keys and publisher URLs in the named key-value store google-news-monitor (entries expire after 45 days). Use monitorStateKey to run several independent monitors.

Legal / responsible use

This Actor returns headlines, links and page metadata only, never full article content. You are responsible for complying with Google's terms (including the terms shown in Google News RSS feeds), publishers' terms and copyright, and the laws that apply to how you store and use the data.