Medium Articles Scraper — Public RSS and article metadata

Collect recent stories from public Medium author, tag, and publication RSS feeds. The actor also accepts direct article URLs and tries to read structured article metadata and public text from each page. It does not log in to Medium.

Use cases

• Track newly published writing by author, topic, or publication.
• Build reading lists and editorial research datasets.
• Supply article feeds to AI agents / MCP workflows.

Input

{
  "authors": [],
  "tags": ["artificial-intelligence"],
  "publications": [],
  "articleUrls": [],
  "maxArticlesPerSource": 5,
  "includeFullText": true
}

Authors accept an @handle or a Medium profile URL. Tags and publications accept a slug or Medium URL. For example, the prefilled input reads five items from the latest AI tag feed. If omitted, maxArticlesPerSource defaults to 50; RSS feeds currently expose about 10 latest items, so increasing the cap cannot add older RSS stories. Direct article URLs are fetched individually.

Output

Charged rows have type: "article"; each row is charged as article-result. Free summary rows report a source's count and status, and free error rows report fetch or parsing failures.

Output fields that may be null

Field	When it is populated	Why it can be null
claps, responses, readingTimeMin	When an accessible article page exposes the value in JSON-LD or Apollo state	RSS does not include these values; a page may be blocked or omit them
isMemberOnly	When page data explicitly identifies member-only or free access	RSS does not report access status; blocked pages and pages without an explicit access marker leave it unknown
text	From <content:encoded> in RSS or a publicly accessible article body, when includeFullText is true	Tag RSS usually has only a short excerpt; a page may be blocked, omit its body, or mark it member-only

Feed article pages are fetched with at most three concurrent requests. A blocked page does not remove the RSS row. Page requests have a 20 second timeout and retry HTTP 429 and 5xx responses with backoff.

Trimmed row from the public tag RSS feed (page-only fields are null when RSS does not expose them):

{"type":"article","url":"https://medium.com/@khalidkhan3398/how-to-use-suno-ai-in-2026-songs-v6-the-new-speech-beta-and-pricing-6fe6069fc0bf","id":"6fe6069fc0bf","title":"How to Use Suno AI in 2026: Songs, v6, the New Speech Beta and Pricing","subtitle":"Learning how to use Suno is the quickest way to turn an idea, a poem or a few lines of lyrics into a finished song, and from this week…","author":{"name":"Kristen Belly ☄️","username":"khalidkhan3398","url":"https://medium.com/@khalidkhan3398"},"publication":null,"publishedAt":"2026-10-04T10:46:23.000Z","updatedAt":"2026-10-04T10:46:23.275Z","tags":["ai","artificial-intelligence","tech","technology","ai-agent"],"claps":null,"responses":null,"readingTimeMin":null,"isMemberOnly":null,"imageUrl":"https://cdn-images-1.medium.com/max/800/1*fOXOrIWD315DfhqJcM3ZFQ.webp","text":null,"source":"tag:artificial-intelligence","scrapedAt":"2026-10-04T11:00:39.917Z"}

Pricing

$1.50 per 1,000 article rows, using the article-result pay-per-event in Apify Console. Summary and error rows are free. Local runs work without pay-per-event configuration.

Limits

Medium RSS currently returns about 10 recent entries per author, tag, or publication feed. The unauthenticated Medium GraphQL endpoint returned HTTP 403 during verification, so archive pagination is not implemented. A live tag feed returned HTTP 200, but the article page returned HTTP 403 with a Cloudflare challenge from this runtime. The blocked response exposed no JSON-LD, Apollo state, or article meta tags, so this runtime could not verify which page fields that live article would otherwise expose. On accessible pages the parser checks JSON-LD, window.__APOLLO_STATE__, and meta tags. RSS content:encoded bodies are converted to plain text when full text is requested. Member-only article text is omitted when page metadata marks it as not freely accessible. Requests time out after 20 seconds, retry HTTP 429 and 5xx up to three times with backoff, and feed article requests run with concurrency up to three.

FAQ

Do I need a Medium account? No. The actor requests public RSS and article pages only.

Why are claps and responses null? RSS does not include these values. They are populated only if an accessible article page exposes them in JSON-LD or Apollo state. A Cloudflare block leaves them null.

Can I scrape more than the latest 10 posts? Not from the public RSS feeds verified here. GraphQL archives were blocked with HTTP 403 and are not used.

Is full text always available? No. Some author and publication feeds include content:encoded; tag feeds often contain only an excerpt. Public page bodies are used when available. Pages can be blocked, and member-only text is excluded.