German News Monitor

Unofficial keyword monitor across ~10 major German news outlets. Not affiliated with, endorsed by, or officially connected to Tagesschau/ARD, Der Spiegel, Die Zeit, FAZ, Süddeutsche Zeitung, Welt/Axel Springer, heise, Handelsblatt, n-tv, or t-online. It reads each outlet's own public RSS feed (or, for Tagesschau, its public JSON API) — the same data anyone gets by subscribing to that feed in a reader.

What is German News Monitor?

A single Apify actor run that pulls the latest headlines from Tagesschau, Der Spiegel, Die Zeit, FAZ, Süddeutsche Zeitung, Welt, heise online, Handelsblatt, n-tv, and t-online, filters them by your keywords, and hands you back a clean, deduplicated JSON dataset — instead of you writing and maintaining ten separate RSS scrapers.

What can German News Monitor do?

• Fetch current headlines from all 10 outlets (or a subset you pick) in one run.
• Filter by keywords: OR-matched, case-insensitive, and umlaut/ß-insensitive — a keyword like Buergermeister also matches Bürgermeister, and strasse matches Straße.
• Exclude items matching a separate exclude-keyword list.
• Optionally fetch each matched article's page and extract its main text (JSON-LD articleBody first, falling back to an <article>/<p> heuristic), flagging truncated: true when a paywall is detected.
• Restrict to items published within a configurable recency window.
• Deduplicate across runs, so a scheduled run only returns what's new.
• Continue gracefully if one outlet's feed is temporarily down — you still get results from the rest, with a clear log entry for the failure.

What data does it extract?

Field	Type	Description
id	string	Deterministic hash of the article URL.
outlet	string	Outlet slug, e.g. spiegel.
outletName	string	Human-readable outlet name, e.g. Der Spiegel.
title	string	Article headline.
summary	string | null	Short summary/teaser from the feed.
text	string | null	Full article text — only populated when fullText input is enabled.
author	string | null	Author name, when the source provides one.
publishedAt	string | null	ISO-8601 publish date.
url	string	Canonical article URL.
imageUrl	string | null	Teaser/hero image URL.
categories	string[]	Section/tag names from the source.
matchedKeywords	string[]	Which of your input keywords matched this item.
language	string	Always "de".
sourceUrl	string	Same as url (kept for pipeline compatibility).
scrapedAt	string	ISO-8601 timestamp this item was scraped.
truncated	boolean	true when full text was requested but is incomplete (e.g. paywall).

Outlets and sources

Outlet	Slug	Source
Tagesschau	tagesschau	Public JSON API: tagesschau.de/api2u/news/ + .../search/
Der Spiegel	spiegel	RSS: spiegel.de/schlagzeilen/index.rss
Die Zeit	zeit	RSS: newsfeed.zeit.de/index
FAZ	faz	RSS: faz.net/rss/aktuell/
Süddeutsche Zeitung	sz	RSS: rss.sueddeutsche.de/rss/Topthemen
Welt	welt	RSS: welt.de/feeds/latest.rss
heise online	heise	Atom: heise.de/rss/heise-atom.xml
Handelsblatt	handelsblatt	RSS: handelsblatt.com/contentexport/feed/schlagzeilen
n-tv	ntv	RSS: n-tv.de/rss
t-online	t-online	RSS: t-online.de/nachrichten/feed.rss

These are the outlets' own public feeds, not a paid data license. Feed availability and structure can change without notice — see NOTES.md for which of these were live-verified versus taken from the outlet's documented feed URL. If a feed 404s or an outlet changes its markup, that outlet is skipped for that run (logged as an error) while the rest of the run continues.

How to use

1. (Optional) Add keywords — e.g. ["Bundestag", "Inflation", "Klimapolitik"]. Leave empty to get everything.
2. (Optional) Pick which outlets to include — all 10 by default.
3. (Optional) Enable fullText if you need full article bodies, not just headlines/summaries.
4. Run the actor. Results land in the run's default dataset.
5. Schedule it (e.g. hourly) with dedupe: true (default) to get a running feed of only new matches.

How much does it cost?

This actor uses the Pay-per-event pricing model — no platform usage charges beyond the events below.

Event	Name	Price	When it's charged
Article delivered	article	$0.001	Once per item pushed to the dataset (after keyword/recency/dedupe filtering).
Full-text extracted	article-fulltext	$0.002	Additionally, only when fullText was enabled and extraction returned text for that item.

Example: a run with fullText: true that returns 100 matched articles, of which 90 yield extractable text, costs 100 × $0.001 + 90 × $0.002 = $0.28. The actor stops pushing further items once your Apify event-charge limit for a run is reached (eventChargeLimitReached), so a run never overspends past your configured cap.

Input

See .actor/input_schema.json for the full schema. Key fields:

{
  "keywords": ["Klimapolitik", "Bundestag"],
  "excludeKeywords": ["Werbung"],
  "outlets": ["tagesschau", "spiegel", "zeit", "faz", "sz", "welt", "heise", "handelsblatt", "ntv", "t-online"],
  "fullText": false,
  "maxItems": 100,
  "maxPerOutlet": 50,
  "publishedWithinHours": 48,
  "dedupe": true
}

An empty input ({}) is valid and runs successfully — it uses every default above.

Output (sample)

{
  "id": "3f2a9c8e1b7d4a56f0e9c1b2",
  "outlet": "spiegel",
  "outletName": "Der Spiegel",
  "title": "Bürgermeister kündigt neue Steuererhöhung an",
  "summary": "Der Bürgermeister der Stadt hat heute eine Steuererhöhung angekündigt.",
  "text": null,
  "author": "Anna Beispiel",
  "publishedAt": "2026-09-02T06:15:00.000Z",
  "url": "https://www.spiegel.de/politik/beispiel-artikel",
  "imageUrl": "https://www.spiegel.de/img/beispiel.jpg",
  "categories": ["Politik", "Lokales"],
  "matchedKeywords": ["Steuer"],
  "language": "de",
  "sourceUrl": "https://www.spiegel.de/politik/beispiel-artikel",
  "scrapedAt": "2026-09-02T09:00:12.345Z",
  "truncated": false
}

FAQ & support

Is this affiliated with any of the news outlets? No. This is an independent, unofficial tool that reads each outlet's own publicly available RSS feed or, for Tagesschau, its public JSON API — the same data available to any RSS reader or app.

Why did an outlet return 0 items in my run? Outlets occasionally change their feed URL or format. Check the run log — a failed outlet is logged with an error and the run continues with the rest. Open an issue if a feed listed above consistently fails.

Does fullText bypass paywalls? No. It extracts whatever text is publicly present in the page HTML (via JSON-LD articleBody or an <article> heuristic). Paywalled articles typically expose only a teaser; those are returned with truncated: true rather than being faked or completed.

Why umlaut-insensitive keyword matching? German text on the web inconsistently uses ü/ö/ä/ß versus their ue/oe/ae/ss transliterations (headline styling, ASCII-only CMSes, etc.). Matching folds both forms so you don't have to enter every keyword twice.

How do I get only new articles on a schedule? Leave dedupe: true (default) and schedule the actor (e.g. hourly). Each run remembers previously-seen article IDs in its key-value store and only returns items not seen before.

I found a bug or a broken feed URL. Please open an issue in the actor's source repository (nerdrx/apify-actors) with the outlet, the run log, and, if possible, the current feed URL from the outlet's own site.