Pinterest Pins Scraper

Fast, low-cost Pinterest pins scraper. One actor, four input modes:

1. Search query: paginate Pinterest's search results for any free-text query.
2. Board URL: every pin from any public board.
3. User profile: every pin from any public user account.
4. Single pin URL: full per-pin detail including the outbound source link.

Direct read of Pinterest's public Resource API with TLS-fingerprint impersonation. No headless browser, no logins, no user cookies. Each pin record is one item in the default dataset.

What you get per pin

Every record has the same shape, regardless of input mode:

• id, url, title, description, altText
• imageUrl (largest) plus allImageUrls (every size variant)
• imageWidth, imageHeight, dominantColor
• isVideo, videoUrl (MP4 for video pins)
• sourceLink, domain: the outbound URL the pin points to
• saveCount, repinCount, commentCount, reactionCount
• createdAt
• pinner: id, username, full name, follower count, profile URL, avatar
• board: id, name, URL, pin count, follower count
• isPromoted, source, sourceInput, scrapedAt

Input

Field	Type	Description
searchQueries	string[]	Pinterest search queries. One pin search per query.
boardUrls	string[]	Pinterest board URLs (e.g. https://www.pinterest.com/{user}/{board}/).
userUrls	string[]	Pinterest user URLs or bare usernames.
pinUrls	string[]	Pinterest pin URLs or bare pin IDs.
maxResults	integer	Cap on pins per search / board / user. Default 100.
enrichWithDetails	boolean	When true, fetch full per-pin detail (adds source link plus aggregated save count).
proxyConfiguration	object	Apify proxy config. Residential US recommended.

You can mix input modes in a single run. Each input is independent; maxResults applies per input, not globally.

Example input

{
    "searchQueries": ["minimalist kitchen", "japandi living room"],
    "boardUrls": ["https://www.pinterest.com/akoerten/scott-lane-kitchen/"],
    "userUrls": ["akoerten"],
    "pinUrls": ["https://www.pinterest.com/pin/4503668374286122/"],
    "maxResults": 50,
    "enrichWithDetails": false
}

Pricing

Pay per result. $0.002 per pin record. No actor-start fee, no rental, no surprise charges.

That is $2.00 per 1,000 pins, undercutting the top Pinterest pin scrapers on Apify by 30 to 50 percent.

How it works

• Cold-start GET to https://www.pinterest.com/ to warm the anonymous cookie session.
• Each input mode hits the matching Pinterest Resource API endpoint:
  • search uses BaseSearchResource
  • board uses BoardResource plus BoardFeedResource
  • user uses UserPinsResource
  • pin uses PinResource
• Pagination is via Pinterest's bookmark token.
• TLS fingerprint impersonation via curl_cffi (chrome124 / chrome131 / safari17_0 rotation) defeats Pinterest's stock-client block at the handshake layer.
• One 403 triggers automatic session rotation and re-warm.

No headless browser. No CDP. No Cloudflare cookie scraping. Just clean JSON API calls.

Output format

The default dataset receives one record per pin in camelCase JSON. Use the Overview view in Apify Console for a quick scrollable table; use ?format=csv on the dataset URL for CSV. See the dataset schema for the full field list and types.

Use cases

• Content marketing: discover trending visuals for any niche.
• Ecommerce trend research: track Pinterest saves on competitor product images.
• Creative inspiration mining: build moodboards programmatically.
• Influencer or pinner discovery: rank users by save velocity.
• Image dataset building for fashion, home decor, food, fitness verticals.

Notes and limits

• Pinterest search re-ranks per session. Two runs of the same query 5 minutes apart return overlapping but different first pages.
• saveCount from search-only results is the per-pin repin count. Use enrichWithDetails: true to get the aggregated cross-pin save count from aggregated_pin_data.aggregated_stats.saves (usually 5 to 50 times larger).
• Pinterest's bookmark cursor walks hundreds of pages deep without re-warming.
• For boards or users with thousands of pins, set maxResults accordingly and budget more memory.
• Residential proxy is recommended for sustained scraping. Datacenter works for small jobs.

Related actors

• Looking for Pinterest follower lists, comments, or board catalogs? Reach out and we will add them.