Article Extractor: News & Blog Text, RSS, $2/1K, No Login

For AI agents: pass article URLs, or RSS, Atom or sitemap feeds; get one JSON row per article with clean text (or Markdown or HTML), title, publication date, site name, language, main image, word count and reading time. Pages that are not articles, and articles older than your date filter, return free rows.

Cost: $0.002 per article extracted ($2 per 1,000). Pages with no article, blocked or missing pages, and feed errors are free. No login or API key.

What does Article Extractor do?

It reads news stories, blog posts, press releases and guides and returns the article itself: the text without menus, ads, related-story lists and comment sections, plus the details a reader or a pipeline needs. You can give it article addresses, or give it RSS, Atom or XML sitemap feeds and it picks the newest articles from each.

• Media monitoring: follow the feeds of news sites, agencies and blogs and get each new article's text once, on a schedule.
• RAG and research: build a clean corpus of articles on a topic, with dates and sources.
• Content teams: read the length, date and summary of many articles at once.
• AI agents: give an agent the readable text of an article through the Apify API or Apify's MCP server.

Try it now. The form opens with a NASA science story and NASA's news feed (3 newest articles). Click Start; it takes a few seconds and costs under a cent.

How to extract articles from URLs or RSS feeds

1. Paste article addresses into Article URLs, or feed and sitemap addresses into Feeds and sitemaps, or both.
2. Set Articles per feed (the newest ones are taken) and, if you want, Published since (a date or a period such as 7 days).
3. Pick the Output formats: plain text, Markdown, clean HTML, or several.
4. For monitoring, turn on Only new articles since my last run, give it a Monitor name, and schedule the Actor. Each run returns only articles it has not returned before.
5. Click Start, then open the Articles or Article text table, or download JSON, CSV or Excel.

How much does article extraction cost?

$0.002 per article extracted ($2 per 1,000 articles). Free rows: pages with no article (home pages, lists, forms), articles older than your date filter, pages that are blocked, missing or not allowed by robots.txt, and feeds that cannot be read. Apify adds only its small standard fee per run start.

Example: you follow 30 news feeds every morning with Only new articles on. About 400 new articles appear each day, so a day costs 400 x $0.002 = $0.80, about $24 a month. You can set a spending limit on any run, and the Actor stops cleanly when it is reached.

Input

{
  "articleUrls": ["https://www.nasa.gov/centers-and-facilities/goddard/decoding-extreme-weather-at-the-poles/"],
  "feedUrls": ["https://www.nasa.gov/news-release/feed/"],
  "maxArticlesPerFeed": 3,
  "publishedSince": "7 days",
  "outputFormats": ["text"],
  "onlyNewArticles": false
}

Advanced options: maximum articles per run (up to 5,000), images kept in the Markdown and HTML, maximum characters per article, articles at a time (at most 2 at a time go to the same site) and a timeout per page.

Output

{
  "url": "https://science.nasa.gov/earth/earth-observatory/uncovering-the-valleys-hidden-below-greenlands-ice/",
  "status": "ok",
  "source": "feed",
  "feedUrl": "https://www.nasa.gov/news-release/feed/",
  "title": "Uncovering the Valleys Hidden Below Greenland’s Ice - NASA Science",
  "siteName": "NASA Science",
  "language": "en-US",
  "publishedAt": "2026-09-28T04:00:00.000Z",
  "imageUrl": "https://assets.science.nasa.gov/...",
  "excerpt": "Greenland is capped with a vast ice sheet that spans 1.7 million square kilometers ...",
  "text": "Greenland is capped with a vast ice sheet ...",
  "wordCount": 688,
  "readingTimeMinutes": 3,
  "truncated": false,
  "charged": true,
  "error": null
}

status is ok for extracted articles, which are charged. The free statuses are not_article, too_old, blocked, robots_disallowed, robots_unreachable, http_error, not_found, timeout, not_html, invalid_url, feed_error and not_a_feed, each with an error that says why. A STATS record in the key-value store counts articles by status, charges, feed items and time.

How does it tell an article from other pages?

A page counts as an article when its main content is long enough to be an article and, for a page given by address, the page marks itself as one (a schema.org Article, NewsArticle or BlogPosting) or states a publication date. Pages from a feed or sitemap are treated as articles when their main content is found. Home pages, section pages and lists come back as free not_article rows with their title.

Related

• Web Page to Markdown: any page, not only articles, as Markdown for LLMs, with JavaScript rendering when a page needs it.

FAQ

Does it respect robots.txt? Yes. It reads each site's robots.txt and skips pages and feeds the site does not allow for the token DonMangu-ArticleExtractor or for all crawlers. Those rows are free.

Can it read paywalled articles? No. It reads public pages only and does not log in. A paywall or login page comes back as blocked (free), and a page that shows only a teaser gives the teaser.

Does it return author names? No. It returns the article and page-level details, not personal details about authors.

Can I reuse the articles? The text belongs to its publishers. Check each site's terms before you republish anything; the Actor is meant for reading, research, monitoring and analysis.

How are sitemaps handled? For a sitemap index, it reads the article sitemaps (news, press, blog and post sitemaps) and skips people, category, page and media sitemaps, then takes the newest entries by date.