Article Text Extractor - News & Blog Content, Date, Image

Paste article URLs, get clean structured data back: title, main text (Markdown or plain), publish date, lead image, site name, language, tags/categories, canonical URL, word count and reading time. Boilerplate (menus, ads, footers, cookie banners) is removed.

Why this one

• Clean text, not page soup. Title, main text (Markdown or plain), publish date, lead image, language and tags in one step, with menus, ads, footers and cookie banners removed.
• The same article is charged once. URLs that differ only by tracking parameters, #anchors, www. or the same canonical URL or identical text are delivered once; the duplicates are not charged.
• Pay only for new articles when monitoring. Set monitorName and run on a schedule: articles already extracted by that monitor are skipped for free.
• Failures are free: blocked, paywalled, non-article or too-short pages cost nothing, and robots.txt is checked for every URL.

Quick start: example input, output and cost

Extract clean article text, publish date, lead image, tags and language from news and blog URLs. Markdown or plain text, no start fee, failures free.

Example input

{
  "urls": [
    {
      "url": "https://www.rfc-editor.org/rfc/rfc9309.html"
    },
    {
      "url": "https://blog.python.org/"
    }
  ]
}

Cost example: 1,000 articles x $0.00099 = $0.99.

Duplicates are charged once, and only new articles on repeat runs

• The same article reached through several URLs (tracking parameters, #anchors, www. or not, the same canonical URL, or identical text) is delivered and charged once. The other URLs appear in ERRORS with "delivered once. Not charged."
• Set monitorName (for example eu-news) and run on a schedule: articles already extracted in an earlier run of that monitor are skipped for free, so you only pay for new ones.

Limits

• It reads the HTML the server sends (no browser), so text, links or markup added later by JavaScript are not seen.
• Paywalled or login-only articles are not extracted (free).
• Pages that are not articles (home pages, listings) are skipped as too short (free).

Pricing (pay per event)

• $0.00099 per article extracted. No start fee.
• Failed, blocked, non-article or too-short pages are free and listed in the ERRORS record with the reason. Your max cost per run is respected exactly.

Good for

RAG / LLM pipelines, news monitoring, content research and summarisation, SEO content audits, building reading lists.

Responsible use

• robots.txt is checked for every URL (RFC 9309) and pages it disallows are skipped (free). Max 1 request per second per site, honest user agent, no browser, no paywall or anti-bot bypass, no logins.
• Social networks and profile sites are refused. Emails and phone numbers in the text are redacted by default. Author bylines are only included if you switch on includeAuthor.
• Respect the copyright of the articles you extract; use the text in line with the publisher's terms.

Maintained by DigiVibeTech.