Linktree Scraper — Bio Links & Destination URLs

Extract every destination URL behind a Linktree bio-link page, plus the account's title, description, avatar, tier, country and creation date. HTTP-only, no account, no browser.

Useful when you have a list of creators or brands and want the links they actually point people at — the shop, the newsletter, the booking page — rather than just the bio-link URL.

Input

Field	Type	Default	Description
profiles	array	required	Usernames (spotify), @handles, or full URLs (https://linktr.ee/spotify)
maxConcurrency	integer	5	Profiles fetched in parallel
proxyConfiguration	object	Apify proxy	Optional — a bare IP worked in testing

Output

One row per input, always — including failures, so downstream joins never silently lose a key.

{
  "_input": "spotify",
  "_source": "S1-next-data",
  "_scrapedAt": "2026-09-07T01:52:39Z",

  "username": "spotify",
  "profileUrl": "https://linktr.ee/spotify",
  "pageTitle": "@spotify",
  "description": "…",
  "profilePictureUrl": "…",
  "tier": "free",
  "country": "US",
  "createdAt": 1557790392,
  "isActive": true,
  "isUnclaimedAccount": false,

  "linkCount": 6,
  "urls": ["https://open.spotify.com/…", "…"],   // flat list of destinations
  "links": [                                      // structured, with Linktree's own fields
    { "id": "9964805", "title": "Playlist 1", "url": "https://open.spotify.com/…",
      "type": "CLASSIC", "position": 0, "_depth": 0 }
  ],
  "pinnedLinks": [],
  "socialLinks": []
}

Failure rows carry _error (not_found, bad_input, blocked, unexpected_shape) and _errorDetail, and no payload.

How it works

Everything comes from the __NEXT_DATA__ island Next.js ships with the page, not from the rendered markup. That is both more stable across redesigns and richer — tier, country, createdAt and link position are all in the island and never shown on the page.

Two details worth knowing, because both cause silent wrong answers if you get them wrong:

Select the island by id, not by a regex over its opening tag. Linktree renders it as <script id="__NEXT_DATA__" type="application/json" crossorigin="anonymous">. A regex written against the usual two-attribute form matches nothing, and the page then looks empty rather than broken.

A 200 is not proof the profile exists. Linktree answers an unregistered handle with a normal 200 "claim this username" page that still carries a __NEXT_DATA__ island — so the parse succeeds and every field comes back null. This actor checks for account identity markers (id, uuid, createdAt, pageTitle, username) and emits not_found instead of a null-filled row that looks like a successful scrape of an empty profile.

Nested links are flattened. Linktree lets a link be an expandable group with children. Each descendant is emitted as its own row in links with a _depth and _parentTitle, so you get every destination URL without walking a tree but can still reconstruct the grouping.

Known limits

Limit	Detail
No click analytics	View and click counts are private to the account owner
No email capture	Linktree's email-signup blocks expose no collected addresses
Gated profiles	Pages behind a password or age gate return no links