Google News Scraper

Google News Scraper turns Google News into clean JSON, CSV or Excel, without login, API key or browser. Search by keyword with every Google News operator, pull the headlines of any topic in 60+ countries, turn news.google.com links into the real publisher URL, and run the trend surge detector: give it terms (or let it pick them from today's headlines) and it tells you which ones are exploding right now compared with their normal daily coverage.

US$ 1.00 per 1,000 articles. Trend rows US$ 0.004 each. No monthly fee, no start fee. Error rows are free.

At a glance

• Price per 1,000 articles: US$ 1.00 (down to US$ 0.85 on Gold and above), no run start fee
• Price per trend term: US$ 0.004 (one term in one country, with its latest articles inside)
• Inputs: Keywords, topics, news.google.com links, trend terms
• Original URL: The publisher link, not the news.google.com redirect (on by default, same price)
• Beyond 100 results: Keyword searches split day by day, so you can go past Google's 100-article feed limit
• Description, image, author: Optional, read from the publisher page, same price
• Error rows: Free, with an errorCode

Real rows

Search flydubai with includeArticleDetails on, collected in the Apify cloud on 2026-09-30 (the Google link is shortened here):

{
  "type": "article",
  "title": "Flydubai Pilot Arrested After 'Apparently' Trying to Crash Israel-Bound Plane, Says Netanyahu",
  "source": "time.com",
  "sourceUrl": "https://time.com",
  "publishedAt": "2026-09-30T16:06:06Z",
  "url": "https://time.com/article/2026/09/30/flydubai-pilot-arrested-crash-attempt-stabbing-cockpit-israel-bound-plane/",
  "googleNewsUrl": "https://news.google.com/rss/articles/CBMirwFBVV95cUxPZHBDSTg5cHQ3VFgyaVFET2poZ1R...",
  "originalUrlResolved": true,
  "description": "The co-pilot of a Flydubai flight from Dubai to Tel Aviv has been arrested after seemingly attempting to crash the plane after stabbing the pilot, Israeli Prime Minister Benjamin Netanyahu said Wednesday.",
  "imageUrl": "https://static.time.com/v3/assets/bltea6093859af6183b/blt7d9889114913cf88/6abd25645336d81e6dedaa76/GettyImages-2297316325.jpg?branch=production&width=6156&quality=75&auto=webp&crop=16:9",
  "author": "Olivia-Anne Cleary",
  "relatedArticles": [],
  "rank": 2,
  "searchQuery": "flydubai",
  "topic": null,
  "country": "US",
  "language": "en",
  "input": "search:flydubai"
}

Trend surge with trendAutoTerms: 5 on the Top stories and Technology headlines (US), same day. The Actor found Tab S12 in the headlines on its own: 222 articles in 24 hours against 15.29 a day the week before. Shown with 1 of its 3 topArticles:

{
  "type": "trend",
  "term": "Tab S12",
  "termSource": "auto:TOP+TECHNOLOGY",
  "country": "US",
  "recentHours": 24,
  "recentCount": 222,
  "recentCountEstimated": true,
  "baselineDays": 7,
  "baselineAvgPerDay": 15.29,
  "expectedInRecentWindow": 15.29,
  "surgeRatio": 14.52,
  "surgeScore": 13.69,
  "isNew": false,
  "saturated": false,
  "dailyCounts": [
    { "date": "2026-09-22", "count": 10, "capped": false },
    { "date": "2026-09-23", "count": 21, "capped": false },
    { "date": "2026-09-24", "count": 18, "capped": false },
    { "date": "2026-09-25", "count": 11, "capped": false },
    { "date": "2026-09-26", "count": 10, "capped": false },
    { "date": "2026-09-27", "count": 10, "capped": false },
    { "date": "2026-09-28", "count": 27, "capped": false }
  ],
  "latestPublishedAt": "2026-09-30T19:53:25Z",
  "topArticles": [
    {
      "title": "Samsung launches Galaxy Tab S12 Ultra and S12+ with Photoshop and a larger Plus screen",
      "source": "tech-ish.com",
      "publishedAt": "2026-09-30T19:53:25Z",
      "url": "https://tech-ish.com/2026/09/30/samsung-launches-galaxy-tab-s12-ultra-and-s12-with-photoshop-and-a-larger-plus-screen/"
    }
  ],
  "googleNewsSearchUrl": "https://news.google.com/search?q=Tab%20S12%20when%3A24h&hl=en-US&gl=US&ceid=US%3Aen"
}

In the same run the ranking was: flydubai 66.96, Tab S12 13.69, Christa Gail Pike 4.39, Xbox 3.97, PS5 3.36, GTA 2.25, Nightmares III 2.13, Asha Sharma 2.11, and openai last with saturated: true (more than 100 articles every single day, so there is no baseline to compare against).

What you can do with it

• Catch stories while they explode: a new AI model, a product launch, a recall, a scandal. The surge detector compares the last hours with the days before and puts the biggest jump on top.
• Media monitoring: your brand, competitors or clients, in any country and language, with the real article link.
• News datasets for AI and research: titles, publishers, dates and links, day by day for as far back as you need.
• SEO and content: what the news is covering on your keywords, and which outlets cover it.

Input

Field	What it does
searchQueries	Keywords. Operators work: "exact phrase", OR, -exclude, site:reuters.com, intitle:
timeRange	any, 1h, 12h, 1d, 7d, 30d, 1y (keyword searches)
dateFrom, dateTo	Optional dates like 2026-09-01 (keyword searches)
topics	TOP (Top stories), WORLD, NATION, BUSINESS, TECHNOLOGY, ENTERTAINMENT, SCIENCE, SPORTS, HEALTH
articleUrls	news.google.com links: an article, a search page, a topic page, or any custom topic or section URL copied from news.google.com (the Actor reads its feed; a link Google has no feed for is a free NOT_FOUND row)
countries	Two-letter codes (US, GB, IN, DE, FR, JP, BR...). Each search, topic and trend term runs once per country
language	Optional; empty = the country's main language
trendTerms	Terms to check for a surge
trendAutoTerms	0 = off. Otherwise how many terms to pick from today's headlines
trendAutoTopics	Where the automatic terms come from (default Top stories)
trendRecentHours	Recent window, default 24
trendBaselineDays	Days before the window that make the baseline, default 7
trendArticlesPerTerm	Newest articles inside each trend row, default 5
maxItems	Per search, topic or link. Default 50; above 100, searches go day by day
onlyNew	Default off. On: each run delivers, and charges, only articles not delivered in previous runs (see "Schedule it daily")
resolveOriginalUrl	Default on: the publisher URL instead of the Google redirect
includeArticleDetails	Default off: description, image and author from the publisher page

{
  "searchQueries": ["\"electric vehicles\" -tesla"],
  "timeRange": "7d",
  "countries": ["US", "GB"],
  "topics": ["TECHNOLOGY"],
  "trendTerms": ["solid state battery", "robotaxi"],
  "trendAutoTerms": 10
}

Schedule it daily: a news alert

Save your searches, topics and countries as a Task, turn on onlyNew and attach a Schedule (daily or hourly). The Actor remembers the articles it already delivered for each search, topic or feed link in each country, in a Key-Value Store in your own account (google-news-scraper-only-new). Every run after the first delivers, and charges, only the articles that were not delivered before; a quiet day ends green with nothing charged. The first run delivers up to maxItems and marks the rest of the current feed as seen, and articles published more than a day before that first run are treated as back catalog, so you never pay for old coverage later. Connect the Task to Slack, e-mail or Google Sheets and new coverage lands where you read it. Delete the store to start over.

{
  "searchQueries": ["\"your brand\"", "your competitor"],
  "countries": ["US", "GB"],
  "timeRange": "1d",
  "maxItems": 50,
  "onlyNew": true
}

Output fields

Articles (type: "article"): title, source (publisher), sourceUrl, publishedAt (ISO, UTC), url (original publisher URL), googleNewsUrl, originalUrlResolved, description, imageUrl, author, relatedArticles (other outlets on the same story, from topic headlines), rank (position in the feed), searchQuery, topic, country, language, input, id, scrapedAt.

Trend rows (type: "trend"): term, termSource (input or auto:<topics>), country, language, recentHours, recentCount, recentCountEstimated, baselineDays, baselineAvgPerDay, baselineCappedDays, expectedInRecentWindow, surgeRatio (recent / expected), surgeScore ((recent + 1) / (expected + 1), the sort key), isNew (no coverage at all before), saturated, dailyCounts, latestPublishedAt, topArticles, googleNewsSearchUrl.

Error codes (never charged)

errorCode	Meaning
NO_RESULTS	Google News returned no new article for this input
NOT_RESOLVED	An article link Google News would not turn into a publisher URL
NOT_FOUND	Google News has no feed at this link
INVALID_URL, INVALID_TOPIC, INVALID_INPUT	A value the Actor does not understand
BLOCKED	Google News did not answer after retries; run again
NOT_REACHED, UNEXPECTED	The run timeout arrived, or something unforeseen broke; only that input is lost

Pricing

Event	Price
Article delivered	US$ 1.00 per 1,000
Trend term analyzed	US$ 0.004 per term per country

1,000 articles cost US$ 1.00 with the original URL included. A trend row replaces 8 or more searches (the recent window plus one per baseline day, each up to 100 articles) and carries its newest articles inside at no extra charge. Duplicates and error rows are free, and there is no Apify compute bill on top.

FAQ

How is this different from a Google News API? Google has no public News API; the "Google News APIs" on the market are third-party services with a monthly plan and a request quota. This Actor reads the public Google News feeds and pages directly, charges per article delivered (US$ 1.00 per 1,000) and adds what those services usually lack: the real publisher URL, day-by-day splitting past 100 results, and the trend surge detector. There is no key to manage; run it on a schedule or through the Apify API.

Why do rows have no description or image, and when should I use includeArticleDetails? The Google News feed carries only the title, the publisher, the date and the link. Description, image and author live on the publisher's page, so the Actor has to open each article to get them. Turn includeArticleDetails on when you need them (same price per article, slower run); leave it off for monitoring, counting and link collection. Some publishers block the page read; those rows keep every other field.

What does saturated mean in a trend row? Google News feeds stop at 100 articles. When every baseline day of a term hits that cap (dailyCounts[].capped all true), the Actor cannot tell what normal coverage looks like, so saturated is true and surgeRatio and surgeScore are null. It happens with broad words like "war" or the biggest brands. Use a narrower term, a quoted name or a product name instead.

How do I schedule monitoring without duplicate articles? Inside a run every article is delivered once (deduplicated by its Google News id). Between runs, turn on onlyNew: the Actor remembers what it already delivered for each search, topic and country and skips it, free, on the next runs. Pair it with a timeRange close to the schedule interval (1h for an hourly run, 1d for a daily one) so each run only reads the new window. Without onlyNew, drop repeats on your side by id, the stable key.

Limitations

• Google News feeds return at most 100 articles. Searches above 100 are split by day; a very busy day still stops at 100 (the log says which days).
• Trend counts are exact below 100 articles per window. When the recent window passes 100, the Actor measures a shorter window and projects it (recentCountEstimated: true); if even one hour passes 100, it reports the floor of 100 per hour. When every baseline day passes 100 (broad words like "war" or big brands), the term is saturated and gets no score: use a narrower term or a quoted name.
• Automatic terms come from capitalized names in headline titles, so they work best in languages with capital letters. They are a starting point: candidates are measured and only the biggest jumps are kept.
• Google News days (dateFrom, dateTo, baseline days) follow Los Angeles time.
• Descriptions and images need includeArticleDetails; some publishers block it and those rows keep every other field.
• Public news pages only. This Actor reads public feeds and pages and respects the sites' terms.

Changelog

• 2026-10-04: onlyNew for scheduled runs: each run delivers, and charges, only articles not delivered before, remembered per search, topic and country in a Key-Value Store in your account.
• 2026-09-30, 0.1: first release. Keyword search with operators, date windows and day-by-day splitting past 100, topic headlines in 60+ countries, news.google.com links, original publisher URLs, optional description and image, trend surge detector with automatic terms, free error rows.