Medium Article & Engagement Scraper

Scrape Medium articles with their clap counts — the number that tells you which posts actually landed. Pull a whole topic, an author's back catalogue, a publication's latest articles with their full text, or a single article end to end.

Why use this actor

• Engagement, not just headlines. Claps and response counts per article, so you can rank a topic by what readers actually responded to rather than by what was published most recently.
• 35 articles per topic in a single request — the richest slice this source gives out.
• Paywall status on every article, so you know which ones your readers can open.
• Full article text where Medium publishes it, in HTML and plain text.
• Author and publication resolved, not left as internal references: name, handle, and the publication each article ran in.
• No account, no login, no API key.
• Honest about what each surface gives. Where a figure is not published, the field is explicitly empty and a flag tells you why — no zeros standing in for missing data.

How it works

1. You pick a mode and give it a topic, author, publication or article URL.
2. The actor reads Medium's public pages and pulls out each article with its engagement figures.
3. Results stream into your dataset — download as JSON, CSV or Excel, or load into a database.

Run it on a schedule against the topics you care about and you have a record of what performs.

Modes

Mode	What you get	Per request
tag	Articles for a topic, with claps	~35 articles
author	An author's recent articles, with claps	~10 articles
publication	A publication's latest articles with the article text (no claps)	10 articles
post	One article with its full text	1 article
reference	Check the source is live and what each surface returns	—

Which modes give claps, and which give text, is a property of the source, not a choice. Topic and author pages publish engagement but only a preview; publication feeds publish the article text but no engagement. Every record's _source field says which you got.

Input

{
  "mode": "tag",
  "tags": ["artificial-intelligence", "programming"],
  "tagSection": "top",
  "maxItems": 200,
  "maxConcurrency": 3,
  "proxyConfiguration": { "useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"] }
}

Field	Type	Description
mode	string	tag (default), author, publication, post or reference.
tag / tags	string / array	Topics. A name, slug or /tag/ URL all work.
tagSection	string	top (~35 articles), archive (newest, ~20) or recommended (~10).
author / authors	string / array	Medium handles or profile URLs, e.g. @medium.
publication / publications	string / array	Publication slugs or URLs, e.g. better-programming.
postUrl / postUrls	string / array	Full Medium article URLs.
maxItems	integer	Maximum records per run. Default 200.
maxConcurrency	integer	Pages fetched in parallel. Default 3.
proxyConfiguration	object	Residential proxy, on by default. Keep it on for scheduled runs.

Output

Real output from mode: "tag":

{
  "postId": "adf87621fb4a",
  "title": "The End of Patterns",
  "subtitle": "On recovery of a practice",
  "url": "https://medium.com/design-bootcamp/the-end-of-patterns-adf87621fb4a",
  "slug": "the-end-of-patterns-adf87621fb4a",
  "clapCount": 77,
  "responseCount": 1,
  "isPaywalled": true,
  "visibility": "LOCKED",
  "isSeries": false,
  "publishedAt": "2026-09-16T07:26:39Z",
  "updatedAt": "2026-09-16T07:26:39Z",
  "publishedAtEpochMs": 1789543599243,
  "authorName": "Joe Alterio",
  "authorUsername": "joealterio_7106",
  "authorId": "14f8b20ac0d7",
  "authorBio": null,
  "authorFollowers": null,
  "publicationName": "Bootcamp",
  "publicationSlug": "design-bootcamp",
  "publicationId": "48e972f5c24e",
  "tags": ["ai", "design", "product-design", "creativity", "patterns"],
  "previewImageAlt": "Antique game of Go Fish, with duos of matching cards.",
  "previewImageId": "1*JRC2jcK0mhRmAprJfaEimw.jpeg",
  "hasFullContent": false,
  "tagQueried": "artificial-intelligence",
  "tagSection": "top",
  "recordType": "POST",
  "_query": "artificial-intelligence",
  "_input": "artificial-intelligence",
  "_source": "medium-tag-apollo-top",
  "_scrapedAt": "2026-09-18T05:32:16Z"
}

Real output from mode: "publication" (content truncated here):

{
  "postId": "3f456a51ff99",
  "title": "Calling AWS Bedrock from code",
  "url": "https://medium.com/better-programming/calling-aws-bedrock-from-code-3f456a51ff99?source=rss----d0b105d10f0a---4",
  "authorName": "Thomas Reid",
  "publishedAtRaw": "Fri, 10 Nov 2023 17:35:02 GMT",
  "updatedAtRaw": "2023-11-10T17:35:02.467Z",
  "tags": ["bedrock", "python", "llm", "boto3", "ai"],
  "contentHtml": "<div class=\"medium-feed-item\"><p class=\"medium-feed-image\"><a href=\"https://medium.com/bet ... (truncated)",
  "contentText": "Using Python in a Jupyter notebook   Continue reading on Better Programming » ... (truncated)",
  "contentSource": "description",
  "isContentFull": false,
  "clapCount": null,
  "responseCount": null,
  "publicationSlug": "better-programming",
  "recordType": "POST",
  "_query": "better-programming",
  "_input": "better-programming",
  "_source": "medium-publication-rss",
  "_scrapedAt": "2026-09-18T05:33:06Z"
}

Field	Type	Description
title / subtitle / url / slug	string	The article and where to read it.
clapCount / responseCount	integer	Engagement. Present in tag, author and post modes; always null in publication mode, because the feed publishes none.
isPaywalled / visibility	boolean / string	Whether the article is member-only.
publishedAt / updatedAt / publishedAtEpochMs	string / number	When it went up, as ISO 8601 and as the source's own epoch milliseconds.
authorName / authorUsername / authorId / authorBio / authorFollowers	string / integer	The author.
publicationName / publicationSlug / publicationId	string	The publication it ran in, if any.
tags	array	Topics the article is filed under.
tagQueried / tagSection / authorQueried	string	What you asked for, kept alongside what came back.
contentHtml / contentText	string	Publication mode: the article body, as HTML and plain text.
contentSource / isContentFull	string / boolean	Whether that body is the full article or just a preview.
bodyText / paragraphs / paragraphCount / isTruncated	string / array / integer / boolean	Post mode: the article text, paragraph by paragraph, and whether the source cut it short.
recordType	string	POST, SURFACE or ERROR.
error / errorDetail	string	Present only on failure rows, so a problem is never a silently missing article.

Limits worth knowing before you run it

• Claps and full text come from different places, and never together. Topic and author pages carry engagement; publication feeds carry article text. If you want both for the same article, run tag to find it and then post to read it.
• No pagination anywhere. A topic gives about 35 articles, an author or publication about 10, and there is no page 2. Cover more ground by querying more topics, more authors and the three different tag sections (top, archive, recommended), which return different sets.
• Member-only articles are truncated at the source — roughly 8 paragraphs instead of 70+. isPaywalled and isTruncated tell you when that has happened, and no attempt is made to get past it.
• A publication's web page cannot be used. It advertises 167 articles but carries no readable article behind any of them, so publication mode reads the feed instead — which is why publication records have no clap counts. reference mode re-checks this every run and will say if it ever changes.
• Publication feeds carry the full article only sometimes. In one measured feed, 4 of 10 articles came through complete and 6 came as short previews. isContentFull tells you which is which per article.