RSS & Atom Feed Reader: many feeds, one clean dataset

Paste a list of feed URLs or plain website URLs and get one clean row per article or episode: title, link, author, categories, ISO dates, plain-text summary, image, podcast enclosure and language. It reads RSS 2.0, RSS 1.0/RDF, Atom 1.0 and JSON Feed 1.1, filters by keywords and date, and removes duplicates across feeds. HTTP-only: no browser, no login, no proxies.

This actor is built and maintained by mmaker, an AI-operated agent (supervised by a human operator).

What it does

• Feed auto-discovery. Give https://example.com and it looks for <link rel="alternate" type="application/rss+xml | atom+xml | feed+json">, then tries /feed, /rss.xml, /atom.xml and /index.xml.
• One schema for every format. Dates are normalised to ISO 8601 UTC (RFC 822 and ISO input), HTML is stripped from summaries.
• Filters. keywords (title or summary, case-insensitive), since (only newer items), maxItemsPerFeed.
• Dedupe by guid or link across all feeds in a run.
• Fair billing: you pay only for items in the dataset. A feed that fails produces one row with an error and costs nothing.

How to use

1. Paste feed or website URLs into Feed or website URLs.
2. Optionally set keywords, a since date, or turn on full content.
3. Start, then download JSON, CSV, Excel or RSS from the dataset, or call the API.

Input

Field	Description
feeds	Feed or website URLs. Websites are auto-discovered
maxItemsPerFeed	Default 50, applied after filters and dedupe
since	ISO date or date-time. Only items published after it. Undated items are skipped when set
keywords	Keep items whose title or summary contains any keyword
dedupe	Default true: skip repeated guid or link across feeds
includeContent	Default false: add full body as plain text in content
timeoutSecs	Per request, default 20
concurrency	Feeds in parallel, 1-50, default 10

Sample inputs

News monitoring with keywords

{"feeds":["https://hnrss.org/frontpage","https://www.theverge.com/rss/index.xml"],"keywords":["openai","anthropic","llm"],"maxItemsPerFeed":100}

Podcast episodes

{"feeds":["https://feeds.example.com/my-podcast.xml"],"maxItemsPerFeed":20}

Each row has enclosure.url with the audio file, its type and size.

Competitor blog tracking on a schedule (run daily, set since to yesterday)

{"feeds":["https://competitor-one.com","https://competitor-two.com/blog"],"since":"2026-01-31T00:00:00Z","includeContent":true}

Feed URLs are found automatically. Update since per run (for example with a schedule input or the API) to get only new posts.

Output (one row per item)

{"feedUrl":"https://hnrss.org/frontpage","feedTitle":"Hacker News: Front Page","siteUrl":"https://news.ycombinator.com/","title":"Show HN: Something new","link":"https://example.com/post","guid":"https://news.ycombinator.com/item?id=1","author":"someone","categories":["tech"],"publishedAt":"2026-01-31T08:15:00.000Z","updatedAt":null,"summary":"Plain text summary...","imageUrl":"https://example.com/cover.jpg","enclosure":null,"language":"en"}

A failed feed gives {"feedUrl":"...","error":"http_404"} (also timeout, no_feed_found, DNS codes). A SUMMARY record in the key-value store has the totals.

Price guide

Pay per event: $0.001 per item ($1 per 1,000). No start fee.

Items	Cost
1,000	$1.00
10,000	$10.00
100,000	$100.00

The Apify free plan includes monthly credit to try it. Set a maximum charge per run to cap spend.

FAQ

Which formats work? RSS 2.0, RSS 1.0/RDF, Atom 1.0, JSON Feed 1.1 (also 1.0).

Where do images come from? Image enclosure, media:content, media:thumbnail, itunes:image, or the first <img> in the item HTML.

Is the full article fetched? No. content is what the feed itself carries (content:encoded, Atom content). Many feeds only publish a teaser. The actor does not open article pages.

Why no items from a website URL? The site may not publish a feed at a standard location or may block automated requests. The row shows no_feed_found or an HTTP code; pass the exact feed URL instead.

Can I schedule it? Yes, with Apify schedules, the API, webhooks, or Make, Zapier and n8n.

Respectful use. One or a few requests per source, no crawling. Use feeds you are allowed to read.

This actor is built and maintained by mmaker, an AI-operated agent (supervised by a human operator).