Google News API — Search News by Keyword, Topic & Country

Google News headlines as clean JSON: search any keyword, read any section (World, Business, Technology, …) or paste a topic or publication link from news.google.com, in any country and language. Every article comes with title, publisher, publish time and — the part other feeds skip — the publisher's real article URL, decoded from Google's redirect link.

Try it in 1 minute: press Start with the prefilled input (two queries, US edition, English). You get up to 100 recent articles per query, each with title, url, source, publishedAt, in about 10–20 seconds. Price: $0.50 per 1,000 articles; queries that return nothing are free.

Use it to:

• Monitor a brand, a competitor, a person in public office or a topic across thousands of publishers at once, in one call.
• Feed an AI agent or a RAG pipeline with fresh headlines and the real article links it can open.
• Build a daily news digest for a country and language Google supports (US, GB, DE, FR, IN, BR, JP, …).
• Pull the headlines of one publication or one Google News topic page by pasting its link.
• Replace a paid news API for headline-level data: no key, no monthly plan, pay per article.

How it works

The Actor reads Google News' public RSS feeds — the same feeds any feed reader can subscribe to — for each query or topic you give, in the edition set by Country and Language. Google does not expose article links directly; its feeds carry news.google.com/rss/articles/… redirects. The Actor decodes them: most are decoded locally from the link itself, the newer ones through Google's own link-resolution endpoint (one small request per article). When a link cannot be decoded, the Google link is kept and urlDecoded is false, so you always have something to open.

No HTML of Google is scraped, no login, no cookies. The first request for each feed goes out directly; the proxy is used only to try again after a network error or a refusal.

Input

Field	Required	What it does
queries	one of queries/topics	Keywords to search, one per line. Google syntax works: "exact phrase", site:reuters.com, after:2026-10-01.
topics	one of queries/topics	Section names TOP, WORLD, NATION, BUSINESS, TECHNOLOGY, ENTERTAINMENT, SCIENCE, SPORTS, HEALTH, or links copied from news.google.com (topic, publication, search).
country	no	Two-letter edition country, default US.
language	no	Edition language, default en (de, fr, es, pt-BR, hi, …).
maxItems	no	Articles per query or topic, default 100 (Google's feed maximum).
publishedSince	no	YYYY-MM-DD: skip older articles (free).
decodeUrls	no	Resolve the publisher's real URL (default on).
maxConcurrency, proxyConfiguration	no	Speed and proxy; the defaults work.

Output

One record per article, plus one free record per query or topic that gave nothing (status NO_RESULTS, NOT_FOUND, INVALID_INPUT or FAILED with an error).

Field	What it is
title	headline, without the trailing publisher name
url	the publisher's article URL (or the Google link when decoding failed)
urlDecoded	true when url is the publisher's address
googleNewsUrl	the original news.google.com redirect link
source, sourceUrl	publisher name and website
publishedAt	ISO 8601 publish time
snippet	short description when the feed carries one (Google News usually does not)
input, query / topic, position	what you gave and the article's place in the feed
country, language	the edition that was searched
status, error, scrapedAt	OK for a delivered article; otherwise why the input gave nothing

Example

{
  "input": "artificial intelligence",
  "query": "artificial intelligence",
  "position": 1,
  "title": "City council approves new AI rules for public offices",
  "url": "https://www.example.com/news/ai-rules-2026",
  "urlDecoded": true,
  "googleNewsUrl": "https://news.google.com/rss/articles/CBMiKmh0dHBzOi8vd3d3LmV4YW1wbGUuY29tL25ld3MvYWktcnVsZXMtMjAyNtIBAA?oc=5",
  "source": "Example Times",
  "sourceUrl": "https://www.example.com",
  "publishedAt": "2026-10-19T09:15:00.000Z",
  "snippet": null,
  "country": "US",
  "language": "en",
  "status": "OK",
  "scrapedAt": "2026-10-19T11:40:12.000Z"
}

Pricing

You pay per article delivered — see the price on this page. Queries and topics without results, invalid inputs and failed fetches are free. Keep the bill predictable with Max articles per query or the run's Maximum charge.

FAQ

Is this an official API? No, it's an unofficial API built on public pages — Google News' public RSS feeds.

How many articles can one query return? Google News caps each feed at about 100 articles. For more, split the search: by after:/before: dates, by site: publisher, or by country edition.

Why is urlDecoded false for some articles? Google's newest redirect links can only be resolved through its link endpoint, which occasionally refuses or times out. The Google link still opens the article in a browser; run again later for the rest.

Does it return the article text? No — headline, source, date and link. Pass the url values to an article extractor to get the full text.

Can I get only new articles on a schedule? Yes: run it hourly or daily as a task with Published since set to yesterday, or plug it into Make, Zapier or n8n. It also works from the Apify API and from AI agents through the Apify MCP server.

Like it? A short review on the Store page helps other people find this Actor. Something missing or broken? Tell us on the Issues tab — we read every one.

Related Actors

• RSS & Atom Feed Reader — any feed or website as one clean list
• Hacker News Scraper — stories, comments and search
• AI Brand Monitor — track your brand in ChatGPT and Perplexity answers
• News Search API — news by keyword with date range and full text
• Article Extractor API — title, author, date and full text of any article