📰 AI News Intelligence Feed — Real-Time News for AI Agents & RAG

The news actor built for AI agents. Turn the world's news into clean, structured JSON — with real publisher URLs, clean full article text, sentiment and de-duplication — ready to feed straight to an LLM, RAG pipeline, or MCP-connected agent. Search Google News (which aggregates Reuters, AP, BBC, Bloomberg, Yahoo, Nature & 1000s more), plus topic sections, top headlines and BBC direct.

✅ No API key, no login, no block. Free public sources. Set it on a daily schedule for an always-fresh news feed.

---

🤖 Why this beats a basic Google News scraper

Most news scrapers just dump headlines with a useless news.google.com/... redirect link. This one gives your AI agent everything it actually needs:

Feature	Basic scrapers	This actor
Headline, source, date, snippet	✅	✅
REAL publisher URL (decoded from Google's encrypted redirect)	❌	✅
Clean full article text (token-efficient, LLM-ready — no HTML)	❌	✅ (optional)
Sentiment score + label per article	❌	✅
De-duplication (same story across 100 outlets → 1)	❌	✅
Topics, top headlines, freshness window, multi-language	partial	✅
Works as an MCP tool for ChatGPT / Claude	—	✅

The real URL matters: without it you can't open, cite, or crawl the article. The clean full text cuts ~94% of the tokens vs raw HTML — so your LLM reads the story, not the page furniture.

---

🎯 Perfect for

• AI agents / assistants — give them current events (LLMs have a knowledge cutoff)
• RAG pipelines — clean, chunk-ready article text with source attribution
• Brand / PR monitoring — track mentions with sentiment
• Finance & crypto — headline sentiment as a signal (bitcoin, a ticker, a company)
• Research & trend tracking — topic sections + freshness windows
• Daily automated feeds — schedule it; get fresh JSON every morning

---

🧾 Input

Field	Description
queries	Search keywords/phrases (Google News aggregates all major publishers)
topics	Topic sections: world, nation, business, technology, entertainment, sports, science, health
includeTopHeadlines	Also pull the main front-page top stories
includeBBC	Also pull BBC News directly (first-party RSS)
freshness	1h / 24h / 7d / 30d — only recent articles (applies to keywords)
maxResults	Total unique articles (after de-dup), 1–2000
resolveUrls	Decode to REAL publisher URLs (recommended, default on)
includeFullText	Add clean, LLM-ready full article text (slower)
includeSentiment	Add sentiment label + score (default on)
language / country	Google News edition (e.g. en / US, hi / IN)

Example input

{
  "queries": ["OpenAI", "bitcoin price"],
  "topics": ["technology", "business"],
  "freshness": "24h",
  "maxResults": 200,
  "resolveUrls": true,
  "includeSentiment": true
}

📤 Output (per article)

{
  "title": "Previewing GPT-5.6 Sol: a next-generation model",
  "source": "OpenAI",
  "source_url": "https://openai.com/index/previewing-gpt-5-6-sol/",
  "google_news_url": "https://news.google.com/rss/articles/CBMi...",
  "url_resolved": true,
  "published_at": "2026-06-26T17:02:16+00:00",
  "snippet": "OpenAI previews its next-generation model...",
  "image": "https://...",
  "sentiment": "positive",
  "sentiment_score": 0.5,
  "query": "OpenAI",
  "topic": null,
  "feed": "google_news_search",
  "language": "en",
  "country": "US",
  "full_text": "…clean article text when includeFullText is on…",
  "scraped_at": "2026-07-03T..."
}

Every run also saves a SUMMARY (totals + source list) in the key-value store.

🤖 For AI agents, MCP & LLMs

This actor is available as a tool for AI agents via the Apify MCP server — ChatGPT, Claude and custom LLM agents can call it directly to fetch current news, resolve real article URLs, read clean full text, and score sentiment in real time. Clean, structured JSON makes it an ideal RAG / tool-use data source for research, monitoring and assistant workflows — closing the LLM knowledge-cutoff gap.

❓ FAQ

Q: Where does the data come from? Google News RSS (which aggregates Reuters, AP, BBC, Bloomberg, Yahoo, Nature and thousands of publishers) + BBC's own RSS. Public, free, no key.

Q: Why "real URLs"? Google News hides every link behind an encrypted redirect. This actor decodes each one to the true publisher URL so you can open, cite and crawl it.

Q: Is the full text really clean? Yes — it's extracted with a precision content extractor (no nav/ads/HTML), so it's compact and LLM-ready.

Q: Can I run it daily? Yes — schedule it in Apify and get a fresh feed automatically.

💵 Pricing (pay per result)

You only pay for what you get — no subscription:

Event	Price
Actor start	$0.01 / run
Per article (real URL + sentiment)	$0.005
Per article with clean full text (LLM-ready)	$0.012

Examples: 500 fresh articles ≈ $2.50 · 500 with full article text ≈ $6.00. The Apify Free plan's monthly credits cover thousands of articles.

🔗 Related Actors (by the same author)

• Lead Enricher Pro — emails, phones, LinkedIn & tech stack from any website
• Google Maps Leads — business leads with emails & phones
• YouTube Transcript Scraper — transcripts/subtitles in 100+ languages
• Google Trends Scraper — keyword & trend data

🔖 Keywords

news api, ai news, real-time news, google news scraper, news for ai agents, llm news feed, rag news, news api for chatgpt, news api for claude, mcp news tool, newsapi alternative, news aggregator, headlines api, article extraction, news sentiment, brand monitoring, crypto news sentiment, reuters ap bbc news, structured news json, daily news feed.

⚖️ Notes

Reads public news headlines/metadata and publicly available article pages. Respect each publisher's terms and copyright when storing or redistributing full text.