Google News RSS - Residential IPs

Pull Google News RSS for any query through a real US residential IP (Verizon AS701) and get back parsed news items plus the raw RSS XML. Each request rotates LRU through a 23-IP residential pool. Every dataset item includes the literal egress IP so you can verify it yourself with ipapi.com or WHOIS.

This Actor is a thin wrapper around the MoneyMaker Residential Scraping API's /v1/scrape/google-news/rss endpoint. Apify hosts and runs the Actor; the residential proxy fleet runs on Hetzner with Verizon AS701 ISP proxies. Apify does not charge for this Actor — you bring your own API key from the MoneyMaker listing on RapidAPI (or directly from the publisher).

---

Why residential matters for Google News

Google News applies the same kind of ASN-level filtering that Search does: high-volume datacenter ranges (AWS, GCP, DO, Azure) get throttled, get served degraded interstitials, or hit a CAPTCHA the moment they cross a volume threshold across the /24. Residential ASNs are treated as ordinary home users.

For the RSS endpoint specifically:

• Datacenter IPs that hammer news.google.com/rss?q=... start hitting a CAPTCHA HTML response in place of XML, often within 50-200 requests.
• A Verizon AS701 IP holds steady for thousands of requests per day at a polite cadence. The 23-IP pool absorbs bursts.

---

Use cases

• News monitoring — daily polling of q="<your brand>" plus q="<your competitors>", push deltas to Slack / Linear / a database.
• Brand-mention tracking — geographic + language coverage by varying gl / ceid (e.g. US:en, GB:en, DE:de, JP:ja).
• AI training pipelines — fresh news headline corpus by topic or domain, Google's de-dup + clustering already applied via the RSS feed.
• Content aggregation — feed a topic newsletter / Discord / RSS reader with curated, source-attributed items.
• Trend analysis — count headline volume by topic-by-day to model attention cycles.

---

Inputs

Field	Type	Required	Description
q	string	yes	Google News query string (OpenAI, site:reuters.com AI)
hl	string	no	UI language tag, default en-US
gl	string	no	Two-letter country code, default US
ceid	string	no	Country-edition pair, default US:en
apiKey	string (secret)	yes	Your MoneyMaker API key (mm_sk_...)
apiBaseUrl	string	no	Default: the public MM origin
verifyTls	boolean	no	Default true

---

Output

Two kinds of dataset items per run.

1. The raw response (one per run)

{
  "kind": "raw_response",
  "ok": true,
  "request_id": "req_w-v8wa1qZ79YcihQ",
  "status_code": 200,
  "ip_used": "65.195.39.208",
  "proxy_node": "Hetzner-12",
  "headers": { "content-type": "application/xml; charset=UTF-8" },
  "body": "<?xml version=\"1.0\" encoding=\"UTF-8\"?><rss>...",
  "body_truncated": false,
  "upstream_latency_ms": 1380,
  "actor_latency_ms": 1450,
  "search": { "q": "OpenAI", "hl": "en-US", "gl": "US", "ceid": "US:en" }
}

2. One parsed news item per dataset item

{
  "kind": "news_item",
  "ok": true,
  "title": "OpenAI announces new pricing tier for enterprise customers",
  "link": "https://news.google.com/rss/articles/CBMi...",
  "pub_date": "Tue, 06 May 2026 14:30:00 GMT",
  "guid": "CBMi...",
  "source_name": "Reuters",
  "source_url": "https://www.reuters.com",
  "description_html": "<ol><li><a href=\"...\">...</a>...</li></ol>",
  "description_text": "...",
  "ip_used": "65.195.39.208",
  "search": { "q": "OpenAI", "hl": "en-US", "gl": "US", "ceid": "US:en" }
}

The Actor's key-value store also gets a SUMMARY record with the search inputs, parsed-item count, the residential IP used, and timing.

---

How it works

Apify run
    │
    ▼
Actor (this code)
    │   GET /v1/scrape/google-news/rss?q=...&hl=...&gl=...&ceid=...
    │   Authorization: Bearer <your MM key>
    ▼
MoneyMaker Scraping API
    │   Picks an LRU residential proxy from the 23-IP pool
    ▼
Verizon AS701 home-broadband IP
    │   GET https://news.google.com/rss/search?...
    ▼
Google returns RSS XML → JSON wrapper to Actor → push_data

You only pay MoneyMaker for the underlying calls — Apify is free. Costs follow the MoneyMaker pricing tiers on RapidAPI.

---

Pacing guidance

• One run is one upstream call. To monitor multiple queries, schedule multiple runs (Apify's scheduler supports cron-like cadence).
• Google News RSS is generally cache-friendly — items don't churn faster than Google's editorial ranking re-ranks (every 5-15 minutes per query). Hourly polling is plenty for most monitoring use cases.
• Cache the guid of items already seen to dedupe across runs.

---

Limitations

• RSS only. No full article body. The link redirects through Google News's URL wrapper to the original publisher; if you need full article text, follow the link with the residential-web-scraper Actor.
• No JS rendering. Static XML only.
• US residential only. All egress is Verizon AS701 / United States. You can still ask Google for any country edition via gl / ceid, but the request itself originates from a US IP.
• No CAPTCHA solving. If Google returns a challenge instead of XML, you'll see the challenge HTML — it's not auto-solved.
• Body cap of 2 MB per response.

---

Source & support

The Actor source lives at sksyen/moneymaker-dashboard/scripts/apify_actor_google_news_rss. Issues and feature requests go on the GitHub repo. Underlying API contract is published as OpenAPI 3.0 at the /openapi.yaml path on the origin.