Thingiverse 3D Model Scraper

Scrape Thingiverse.com — the world's largest 3D-printing design repository, home to millions of free, downloadable STL/3MF files. Look up a design by thing ID or share URL, browse the newest published designs, fetch a maker/uploader's profile, or keyword-search the whole catalog. HTTP-only. No auth, no login, no cookies.

What this actor does

• Five modes: byIds, byUrls, latest, byUser, search
• Full model metadata: title, description, category, license, images, likes, comments, publish/update dates
• Maker profiles: display name, bio, avatar, user ID
• Filters: license type, minimum like count
• Empty fields are omitted — a field only appears on a record when Thingiverse actually published a value for it

Output per thing (modes byIds, byUrls, latest, search)

• thingId — numeric Thingiverse thing ID
• title, description
• category, subCategory — e.g. 3D Printing / Toys & Games
• imageUrl — full-resolution primary image
• authorName, authorUsername, authorUrl — the maker who uploaded the design
• license — friendly license name (e.g. Attribution (CC BY), Public Domain (CC0))
• isFree — whether the design is free to download
• likesCount, commentsCount
• datePublished, dateModified
• commentsUrl — link to the design's comment thread
• thingUrl — canonical share URL
• recordType: "thing", scrapedAt

Output per maker profile (mode byUser)

• username — canonical Thingiverse handle
• displayName — the maker's public display name
• alternateName — shown only when it differs from username
• userId — numeric Thingiverse user ID
• bio
• avatarUrl
• profileUrl
• recordType: "maker", scrapedAt

Input

Field	Type	Default	Description
mode	string	byIds	byIds / byUrls / latest / byUser / search
thingIds	array	["763622"]	Numeric thing IDs to look up (mode=byIds)
thingUrls	array	–	Thing share URLs to look up (mode=byUrls)
usernames	array	–	Thingiverse usernames or profile URLs (mode=byUser)
searchQuery	string	mount	Keyword matched against recently published things (mode=search)
license	string	any	Only emit things under this license (cc0, cc-by, cc-by-sa, cc-by-nd, cc-by-nc, cc-by-nc-sa, cc-by-nc-nd, gpl, lgpl, bsd, other)
minLikes	int	–	Drop things with fewer likes than this
maxItems	int	20	Hard cap on emitted records (1–100)

Example: batch lookup by thing ID

{
  "mode": "byIds",
  "thingIds": ["763622", "3717484", "2278"]
}

Example: lookup by share URL

{
  "mode": "byUrls",
  "thingUrls": ["https://www.thingiverse.com/thing:763622"]
}

Example: browse the newest published designs

{
  "mode": "latest",
  "maxItems": 50
}

Example: a maker's profile

{
  "mode": "byUser",
  "usernames": ["CreativeTools", "MakerBot"]
}

Example: keyword search filtered to public-domain designs

{
  "mode": "search",
  "searchQuery": "mount",
  "license": "cc0",
  "maxItems": 30
}

Use cases

• 3D printing marketplaces — enrich a product catalog with real Thingiverse metadata and licensing
• Design discovery tools — surface newly published models for a "what's new" feed
• License compliance — filter a batch of designs down to only CC0/public-domain or commercial-safe licenses
• Maker analytics — track a designer's published work, likes, and comment engagement over time
• Academic/dataset research — bulk-export bibliographic-style metadata for 3D model corpora

Limitations

• Large maxItems runs take longer. Thingiverse rate-limits individual thing/profile page requests per source IP much more aggressively than its static sitemap files. Keep maxItems modest (≤50) for fast, predictable runs, and budget more time for larger batches.
• mode=search with a rare/absent keyword still has to scan its whole candidate window before it can report zero matches — increase the run's timeout (Run options → Timeout) beyond the default if you search for uncommon terms with a larger maxItems.
• No tag/category browse mode. Thingiverse's /tag:<name> and category-browse pages don't expose server-rendered listing data (unlike individual thing/profile pages). Use the license filter plus mode=search's category/subCategory keyword matching as a partial substitute.
• No trending/popular mode. Thingiverse's /popular and /newest listing pages have no public, unauthenticated data feed. mode=latest (sourced from the public sitemap) is the closest reliable substitute for "what's new."
• mode=search is a recent-catalog keyword filter, not a full-text index search — see the FAQ below and the input field description.

FAQ

What is Thingiverse? A free community platform, owned by MakerBot/Formlabs, where makers upload and share digital designs for 3D printing, laser cutting, and other digital fabrication.

Is this affiliated with Thingiverse? No, this is a third-party actor that reads Thingiverse's own publicly published page data. It is not endorsed by or affiliated with Thingiverse, MakerBot, or Formlabs.

Are all Thingiverse designs free? Yes — every thing on Thingiverse is free to download. The license field tells you the terms under which you may reuse or remix the design (attribution required, non-commercial only, share-alike, or fully public domain).

How does mode=search work, and why does it sometimes return fewer results than maxItems? Thingiverse has no public full-text search API for its full catalog. So search scans the most recently published things (via the same public sitemap mode=latest uses) and keyword-matches each one's title/description/category. This means very specific or rare keywords may return fewer results, or none, if no match appears in the scanned window — try a more common term, or use mode=latest for guaranteed-fresh designs regardless of keyword. Every returned thing is a genuine, live Thingiverse page — results are just bounded to a recent-catalog window rather than an all-time index.

Why is there only one imageUrl per thing? Thingiverse's publicly published page data exposes one representative image per design. The full image gallery lives behind the site's interactive app and isn't part of the public page data.

What does mode=latest actually return? The most recently published things, sourced from Thingiverse's own public sitemap (the same file search engines use to discover new designs), each hydrated with full metadata.

How fresh is the data? Every run fetches live pages at request time — nothing is cached or pre-scraped.

Why are tags not included? Thingiverse only exposes per-thing tags through its interactive, login-aware web app, not on the publicly published page — including a fabricated or partial tag list would be worse than omitting it.