Tumblr Posts Scraper

Scrape public posts from any Tumblr blog, tag, or search page. HTTP-only — no cookies, no login, no API key input, no proxy required.

What this actor does

Given a list of Tumblr URLs or a tag query, this actor extracts public post data including:

• Blog name, title, description, URL
• Post ID, URL, slug, post type
• Publish date (ISO-8601) and Unix timestamp
• Tags
• Note count (engagement)
• Content text (HTML stripped)
• Per-type fields: title + body (text), image URLs (photo), video source (video), quote text + source (quote), link URL (link), audio player (audio)
• Scrape timestamp

Input

Field	Type	Required	Description
startUrls	string[]	At least one input required	Any mix of: https://<blog>.tumblr.com, https://www.tumblr.com/<blog>, https://www.tumblr.com/tagged/<tag>, https://www.tumblr.com/search/<query>, https://<blog>.tumblr.com/tagged/<tag>.
searchQuery	string	Optional	A tag or search term. Uses Tumblr's public tagged feed. Can be combined with startUrls.
postType	enum	No (default any)	Filter by post type: any, text, photo, video, quote, link, audio. Applies to blog URL scrapes.
maxItems	integer	No (default 50)	Maximum posts across all inputs (1-1000).
country	enum	No (default us)	Accept-Language hint (us, gb, de, jp).

You must provide at least one of startUrls or searchQuery.

Output

Every record is one post with at minimum:

• type: always "tumblr_post"
• blogName, blogUrl
• postId, postUrl, postType
• date (ISO-8601), timestamp
• tags (array, may be empty)
• noteCount (integer)
• contentText (stripped HTML, may be empty string for media-only posts)
• scrapedAt

Type-specific fields (emitted only when present): title, bodyHtml, imageUrls, photoLinkUrl, videoSource, videoPlayer, quoteText, quoteSource, linkUrl, linkText, audioPlayer, question.

How it works (high level)

• Blog URLs hit Tumblr's legacy /api/read/json endpoint directly on the blog subdomain. No API key, no rate limit issues.
• Tag / search queries use Tumblr's public api.tumblr.com/v2/tagged endpoint with the official Tumblr demo API key.
• Requests impersonate Chrome TLS fingerprint via curl_cffi to avoid false-positive anti-bot blocks.

FAQ

Do I need a Tumblr API key? No. The actor uses Tumblr's own published demo key, which is rate-limited but sufficient for typical search usage. For blog URLs the key is not used at all.

Do I need a proxy? No. Tumblr does not block datacenter IPs for these endpoints.

Why does maxItems sometimes return fewer posts? Some blogs have fewer posts than the requested limit, or the tag search hits its server-side cap before reaching your limit.

Custom-domain blogs (e.g. humansofnewyork.com) are not supported — Tumblr does not route their URL through the standard subdomain API. Use the underlying <blog>.tumblr.com form.