📰 Google News Extractor — No Login, Full Text & Bulk Search

Extract structured news article data from Google News by keyword, topic, or URL — no login, API key, cookies, or browser required. Built for developers, data teams, AI/LLM pipelines, and automated news monitoring workflows.

Runs on Apify — schedule it, call it via API, chain it with other actors, or integrate it directly into your MCP-compatible AI agent.

---

What It Does

This actor fetches Google News RSS feeds, parses them into clean structured records, and optionally visits article pages to extract full body text. Each run produces a consistent JSON dataset you can export to CSV, Excel, JSON, or push to any downstream system.

Output fields per article: title, source_name, source_url, google_news_url, published_at, description, image_url, search_query, topic, full_text, word_count, scraped_at

---

Key Features

No login or API key required — works out of the box with no authentication setup.

Bulk keyword search — pass multiple search queries in one run. Each query gets its own RSS feed and result set. Run 50 queries in a single call.

Built-in topic sections — monitor World, Technology, Business, Science, Health, Sports, Entertainment, or Nation headlines with a single click.

50+ region/language editions — scrape Google News in any edition: US English, German, French, Japanese, Arabic, Spanish, and 45+ more. Every language correctly sets hl, gl, and ceid parameters.

Google search operators supported — use quotes for exact phrases, minus to exclude terms, OR for alternatives, and site: to filter by publisher:

Operator	Example	Effect
"exact phrase"	"interest rate hike"	Match exact wording
-keyword	apple -fruit	Exclude term
OR	AI OR "machine learning"	Either term
site:	site:reuters.com earnings	Publisher filter

Time range filtering — filter by past hour, 24 hours, 7 days, 30 days, or 1 year.

Full article text extraction — optionally visit each article URL and extract the full body text. Produces full_text and word_count fields. Designed for AI/RAG pipelines, sentiment analysis, and NLP workloads.

Deduplication — across multiple queries and topics, duplicate articles are removed automatically.

Custom topic URLs — paste any Google News section URL to extract niche topic feeds not covered by built-in sections.

Structured output schema — fully defined dataset schema with typed fields, descriptions, and examples. MCP-compatible: field names and descriptions are optimized for LLM tool-use routing.

---

Input Options

Field	Description
searchQueries	Array of keywords/phrases. Supports Google operators.
topics	Built-in sections: WORLD, TECHNOLOGY, BUSINESS, etc.
topicUrls	Custom Google News section URLs (HTML or RSS format).
startUrls	Raw RSS feed URLs for advanced use.
maxResultsPerQuery	1–100 articles per query/topic (default: 100).
regionLanguage	Edition code like US:en, DE:de, JP:ja (50+ options).
timeRange	1h, 1d, 7d, 30d, 1y, or any.
extractFullText	Visit article pages and extract body text.
decodeUrls	Attempt to resolve real article URLs from Google redirects.
deduplicateResults	Remove duplicate articles across queries (default: true).

---

Output Schema

Each article record has these fields:

Field	Type	Description
title	string	Article headline, cleaned (source name suffix removed)
source_name	string	Publisher name (e.g. "Reuters", "BBC News")
source_url	string|null	Real article URL (when resolvable)
google_news_url	string	Google News redirect URL (always present)
published_at	string	ISO 8601 publication datetime
description	string|null	Article snippet/summary
image_url	string|null	Article thumbnail image URL
search_query	string|null	The query that found this article
topic	string|null	Built-in topic section (if applicable)
full_text	string|null	Full article body text (requires extractFullText)
word_count	integer|null	Word count of full text (requires extractFullText)
scraped_at	string	ISO 8601 extraction timestamp

---

Usage Examples

Monitor a brand or topic daily

{
  "searchQueries": ["OpenAI", "Anthropic", "Google DeepMind"],
  "maxResultsPerQuery": 20,
  "timeRange": "1d",
  "deduplicateResults": true
}

Collect top headlines across sections

{
  "topics": ["TECHNOLOGY", "BUSINESS", "SCIENCE", "HEALTH"],
  "maxResultsPerQuery": 10
}

AI/RAG pipeline — full text extraction

{
  "searchQueries": ["large language models", "AI regulation EU"],
  "maxResultsPerQuery": 15,
  "extractFullText": true,
  "timeRange": "7d"
}

Monitor non-English news

{
  "searchQueries": ["Künstliche Intelligenz", "Bundesliga"],
  "regionLanguage": "DE:de",
  "maxResultsPerQuery": 25
}

Bulk competitive intelligence

{
  "searchQueries": [
    "Tesla earnings", "Ford EV", "GM electric",
    "Rivian news", "Lucid Motors", "NIO stock",
    "BYD sales", "Volkswagen EV", "BMW electric", "Mercedes EV"
  ],
  "maxResultsPerQuery": 10,
  "timeRange": "7d",
  "deduplicateResults": true
}

---

Performance & Cost

Mode	Articles/min	Cost estimate
Metadata only	~200	$0.003 per article
With full text	~30–60	$0.003 + $0.005 per article

Example costs:

• 100 articles (metadata): $0.30
• 100 articles (with full text): $0.80
• 1,000 articles (metadata only): $3.00
• Daily monitoring, 10 queries × 20 articles: $0.60/day

Google News RSS returns up to 100 articles per feed. A single run with 10 search queries can collect up to 1,000 articles.

---

API & MCP Integration

Call via Apify API

curl -X POST "https://api.apify.com/v2/acts/khadinakbar~google-news-scraper/runs?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"searchQueries": ["artificial intelligence"], "maxResultsPerQuery": 10}'

Use in an AI agent (MCP)

This actor is optimized for use via the Apify MCP server. When an LLM agent calls call-actor with khadinakbar/google-news-scraper, the structured input schema and output schema enable precise tool-use routing without hallucination.

Typical LLM agent prompt:

"Get the latest 10 news articles about AI regulation from the past week using Google News"

Agent call:

{
  "searchQueries": ["AI regulation"],
  "maxResultsPerQuery": 10,
  "timeRange": "7d"
}

---

Comparison vs. Competitors

Feature	This Actor	data_xplorer	easyapi	scrapestorm
Output schema (typed)	✅	❌	❌	❌
MCP-optimized fields	✅	❌	❌	❌
Bulk keywords (1 run)	✅	✅	❌	✅
Custom topic URLs	✅	❌	❌	❌
Full text extraction	✅	❌	❌	❌
Deduplication	✅	❌	❌	❌
50+ regions	✅	✅	✅	✅
Dataset schema	✅	❌	❌	❌
Limited permissions	✅	❌	❌	❌

---

Notes

• source_url may be null for some articles. Google News encodes article URLs using a Base64 scheme that requires JavaScript to decode. The decodeUrls option attempts HTTP redirect following but cannot resolve all URLs. The google_news_url field is always populated and can be used to visit the article in a browser.
• full_text requires source_url to be resolvable. Articles where source_url is null will have full_text: null even when extractFullText is enabled.
• Rate limits: Google News RSS feeds are public and rate-limit tolerant. This actor runs within safe request limits.
• No blocked content: This actor only fetches publicly available RSS data and article pages. It does not bypass paywalls.

---

Support

Found an issue or want a feature? Open a request via the Issues tab on the actor page.