Fast RSS & Atom Feed Reader — Bulk Feeds, OPML, Media URLs

Point it at any number of RSS 2.0, Atom or RDF feeds and get back one flat, normalized row per item — the same shape no matter which format the publisher used.

No API key. No browser. No proxy. A run over a handful of feeds finishes in seconds.

What you get, per item

Field	Notes
feedUrl, feedTitle, feedType, language	feedType is rss, atom or rdf
id, title, link	id falls back to the link when the feed omits a GUID
publishedAt, updatedAt	ISO 8601 UTC, normalized from the five date formats feeds use in practice
summary	as published
categories	array of tags/categories
imageUrl	Media RSS thumbnail, else the feed's own image
enclosureUrl, enclosureType, enclosureLength	the podcast MP3, the video, the attachment
content	optional; see below

A per-run FEED_SUMMARY record in the key-value store lists every feed with its item count, so you can tell an empty feed from a failed one.

Input

{
  "feedUrls": [
    "https://feeds.bbci.co.uk/news/rss.xml",
    "https://techcrunch.com/feed/"
  ],
  "maxItemsPerFeed": 50,
  "sinceDaysAgo": 7,
  "includeFullContent": false
}

• opmlUrl — point it at an OPML file (every feed reader exports one) and every feed inside is added to the run. Useful for migrating a whole subscription list.
• sinceDaysAgo — skip older items. Items with no date are always kept.
• includeFullContent — adds content, taken from content:encoded in the feed itself. Free, and it can't fail.

Run it with empty input and it reads three public feeds so you can see the output shape before wiring anything up.

What this Actor deliberately does not do

It never fetches the article page. Plenty of feeds ship only a summary, and it is tempting to go download each article for the full text. That is exactly where scrapers break: publishers put the feed in the open and the article behind a bot wall. A field that works four times out of five is worse than a field that isn't there.

It does not collect people. Feeds carry author and dc:creator, usually a person's name. Those fields are not mapped. A feed publishes stories, not people.

Known limits, measured

• Reddit feeds (reddit.com/r/*/.rss) return HTTP 403 from Apify's network, with or without proxy. Reddit now gates its feeds. Not supported, and not billed — a feed that fails writes no items.
• A feed that is unreachable or malformed is logged and listed in FEED_SUMMARY with ok: false. You are never charged for it.
• Feeds are fetched 5 at a time to stay polite with publishers.

Verified on 2026-08-04 from Apify's own infrastructure: NYT, BBC, Google News, Hacker News, TechCrunch, WordPress, Substack, YouTube channel feeds, arXiv and Clarín all returned 200 on every attempt, with and without proxy.

Pricing

Pay per event: a small charge when the run starts, plus one charge per feed item returned. Failed feeds return nothing and cost nothing.