RSS Feed Scraper: RSS, Atom & JSON Feed reader to JSON / CSV

Turn any RSS, Atom or JSON Feed into clean, structured data: title, link, publication date, author, categories, summary, full article text (when the feed has it), image and podcast/video enclosures. Paste feed URLs or just website URLs: the feed is discovered automatically. Read hundreds of feeds in one run, filter by keywords and date, remove duplicates across feeds, and schedule it as a news or content monitor. You pay only for the items you get.

Why this one: $0.30 per 1,000 items, while other RSS actors on the Store charge $1–8 per 1,000 items or $0.035 per run plus per-item fees. It is a plain, fast HTTP reader (no browser), handles RSS 2.0, RSS 1.0 (RDF), Atom and JSON Feed, and feeds that fail, filtered-out items, duplicates and already-seen items are free.

Why this scraper

• Every feed format. RSS 2.0, RSS 1.0/RDF, Atom 1.0 and JSON Feed 1.x, all normalized to one schema.
• Feed auto-discovery. Give techcrunch.com or a blog URL; the Actor finds the feed via <link rel="alternate"> or common paths (/feed, /rss.xml, /atom.xml, /index.xml...). Optionally read all feeds a site lists.
• Full content and media. content:encoded / Atom content as plain text (and optionally HTML), image from media tags or the content, enclosures with URL, type and size (podcast episodes, videos).
• Filters. Keywords (any), exclude keywords, published within N days, max items per feed.
• Dedupe + monitoring. Duplicates across feeds are skipped; with Only new items each scheduled run returns only items you have not seen yet.
• Bulk and fast. 100 feeds in about 25 seconds at 256 MB, with a per-feed report of what worked and what failed.

What you get

Field	Example
feedUrl, feedTitle, feedLink, feedType, language	https://techcrunch.com/feed/, TechCrunch, https://techcrunch.com/, rss, en-US
itemId, title, link	GUID, OpenAI launches ..., https://techcrunch.com/2026/10/08/...
publishedAt, updatedAt	2026-10-08T08:29:14.000Z
author, categories	Jane Doe, [AI, Startups]
summary, contentText, contentHtml	plain-text summary, full text, optional HTML
image, enclosures	image URL, [{url: .../episode.mp3, type: audio/mpeg, length: 123}]

How to use

1. Add feeds or websites (one per line), e.g. https://www.nasa.gov/feed/, blog.cloudflare.com, https://github.com/apify/crawlee/releases.atom, a YouTube channel feed or a podcast RSS feed.
2. Optional: keywords, exclude keywords, published within N days, max items per feed.
3. Run, then download the dataset as JSON, CSV, Excel or HTML, or use the API, webhooks or integrations (Google Sheets, Make, Zapier, n8n, Slack).
4. For monitoring, turn on Only new items and add a schedule (e.g. every hour).

Example input

{
  "feeds": ["https://www.nasa.gov/feed/", "blog.cloudflare.com", "https://github.com/apify/crawlee/releases.atom"],
  "keywords": ["ai", "openai", "nvidia"],
  "publishedWithinDays": 2,
  "maxItemsPerFeed": 50,
  "includeContent": true
}

Example output (shortened)

{
  "feedUrl": "https://techcrunch.com/feed/",
  "feedTitle": "TechCrunch",
  "feedType": "rss",
  "title": "OpenAI launches a new model for developers",
  "link": "https://techcrunch.com/2026/10/08/openai-launches-a-new-model/",
  "publishedAt": "2026-10-08T08:29:14.000Z",
  "author": "Jane Doe",
  "categories": ["AI", "Startups"],
  "summary": "The company said ...",
  "contentText": "Full article text ...",
  "image": "https://techcrunch.com/wp-content/uploads/2026/10/image.jpg",
  "enclosures": []
}

Common use cases

• News and brand monitoring: hourly runs over 50 news feeds with keywords, sent to Slack or a Google Sheet.
• Content aggregation: build a niche newsletter or website feed from many blogs, deduplicated.
• AI and RAG pipelines: full-text items from trusted sources as clean JSON for LLM summarization and embeddings.
• Podcast and YouTube tracking: new episodes and videos via their official RSS feeds, with enclosure URLs.

Run report

• FEEDS: for every input: the resolved feed URL, its title, items fetched and saved, and any error (404, blocked, not a feed).
• SUMMARY: totals for the run (feeds read and failed, items fetched, filtered, duplicates, already seen, saved).

Pricing

Pay per event:

• Item (item): one charge per feed item saved to your dataset (full text included).
• Free: feeds that fail or are not found, items removed by your filters, duplicates and already-seen items in monitoring mode.

See the Pricing tab for the current price. Set a maximum cost per run and the Actor stops gracefully when it is reached.

Limitations

• The Actor returns what the feed publishes. Many feeds contain only the latest 10–50 items and a short summary; it does not open article pages to extract text (use a content extractor for that).
• Some sites block datacenter IPs or require a login for their feeds; these show up as errors in FEEDS (and are free). A proxy can help for blocked feeds.
• Email addresses in author fields are removed.

Use with AI agents (MCP)

This Actor works as a tool for AI agents through the Apify MCP server at https://mcp.apify.com. Connect Claude, Cursor, VS Code, n8n or any other MCP client to https://mcp.apify.com?tools=tinyrex/rss-feed-scraper and the agent can run it from a plain-language request, for example: "Read these 20 company blogs and give me this week's posts about AI with links". Results come back as clean, structured JSON at the same pay-per-result price, and filtered or duplicate items stay free.

📘 Step-by-step guide with Python, JavaScript, curl and MCP examples: How to turn any RSS or Atom feed into JSON or CSV (bulk, with monitoring)

Related actors

• Tech Stack Detector: Find which CMS, analytics and email provider the sites behind your feeds use.
• Remote Jobs Scraper: Remote jobs from Himalayas, Remote OK, We Work Remotely and more.
• Sitemap Scraper: Every URL from a website's XML sitemaps, with lastmod and an optional broken link check.
• TED Tenders Scraper: Daily EU public tender alerts from the official TED API.

FAQ

Is it legal? RSS, Atom and JSON feeds are published by sites specifically for syndication and automated reading. The Actor only reads the feeds you give it (and the page needed to discover them), does not log in and does not bypass paywalls. You are responsible for respecting each publisher's terms and copyright: some feeds (e.g. many news outlets) allow only personal, non-commercial use, so check before republishing or using content commercially.

Can it read Google News or social media feeds? It reads any standard feed URL you provide, but some services restrict commercial use of their feeds; check their terms first.

How fast is it? About 100 feeds in 25 seconds with the default settings.