Google Alerts emails are hard to automate, and the Google News RSS feed gives you news.google.com/rss/articles/... redirect links instead of the publisher's URL. This tutorial builds a news monitor that:
- searches Google News for any keywords, sites, language and country
- returns the publisher's real article URL, publisher, date and related coverage
- returns only articles you haven't seen yet, on a schedule
- delivers them to Slack, email or a spreadsheet
One article looks like this
json
{
"title": "Trump AI deal rebrands 'artificial intelligence' as 'superintelligence'",
"url": "https://www.theguardian.com/us-news/2026/sep/29/trump-ai-deal-tech-ceos-superintelligence",
"publisher": "theguardian.com",
"publishedAt": "2026-09-30T17:37:00+00:00",
"query": "artificial intelligence",
"language": "en",
"country": "US",
"urlResolved": true
}Step 1 - choose what to watch
- Search queries take Google News operators:
"exact phrase",-exclude,OR,intitle:word. Example:"Acme Corp" -stock. - Topics: Top stories, Business, Technology, Science...
- Only from these sites / Never from these sites: e.g.
reuters.com,bbc.co.uk. - Published within: past hour, 24 h, week...
- Language / Country:
de+DE,hi+IN,ja+JP, and so on.
Step 2 - try it from code
python
import os, requests
articles = requests.post(
"https://api.apify.com/v2/acts/tacps126~google-news/run-sync-get-dataset-items",
headers={"Authorization": f"Bearer {os.environ['APIFY_TOKEN']}"},
json={"queries": ['"generative AI" -stock'], "timeframe": "1d", "maxItems": 50},
timeout=300,
).json()
for a in articles:
print(a["publishedAt"][:10], a["publisher"], "-", a["title"])
print(" ", a["url"])In my test runs, 100 articles come back in about 11 seconds, publisher URLs included.
Step 3 - turn it into alerts
- Save the input as a task and turn on Only new articles. Each run then returns only articles no earlier run delivered, and you pay only for those.
- Add a schedule: hourly for breaking news, daily for a digest.
- Add an integration: Slack (a message per run), Google Sheets (append rows), email, or a webhook into your own app or RAG index.
Running several monitors? Give each a different Monitor name so they keep separate histories.
Use cases
- PR and brand monitoring: your brand, executives, competitors.
- Investment research: tickers and company names, filtered to trusted publishers.
- AI and RAG pipelines: a fresh, deduplicated stream of article URLs to fetch and embed.
- Local news in any language, using the country and language editions.
Cost
About $3 per 1,000 articles, and with "only new" you pay only for articles you haven't already received.
Missing a filter? Open an issue on the Actor's Issues tab.