This changelog highlights releases that materially changed Pinterest Scraper for users. Dates use UTC; small internal refactors and routine maintenance are intentionally omitted.
Feature
- Video and idea pins now include
videoUrl (public HLS stream), videoDurationSec, and videoThumbnailUrl.
isVideo now detects regular video pins in search, board, and profile results; previously only idea-pin videos were flagged.- Profile pins now report the pinner's follower count when the pinner is the scraped profile, instead of
0.
- Runs now end with a short status message: the number of unique pins and videos, failed inputs, or that no pins were found.
Reliability
- Pin URLs that include Pinterest's title slug (
https://www.pinterest.com/pin/recipe-title--123456789/) are now accepted instead of being rejected as invalid.
Reliability
- Board section URLs (
https://www.pinterest.com/username/board-name/section-name/) now return that section's pins instead of failing.
Reliability
- Preserved recoverable embedded pins and image metadata after malformed or nested noscript fallback markup.
- Reduced transient memory used to read session pages and individual pins, preserving embedded pin data, image metadata and fresh-route recovery without building full page DOM trees.
Reliability
- Board share links and redirected board URLs now use the resolved board path for pin extraction, while preserving the original input URL in result provenance.
Reliability
- Faster result persistence for large requests reduces time spent saving pins near the run timeout, while preserving deduplication, source provenance and per-pin charging.
Reliability
- A failing or invalid profile in a mixed request no longer discards useful pins from other searches, boards, or pin URLs. Requests with no pins still fail clearly.
Reliability
- Shortened stalled upstream requests and removed nested same-route retries so fresh proxy recovery occurs before the run timeout.
Reliability
- Added bounded fresh-session and proxy-identity recovery when Pinterest intermittently returns an empty response or an upstream proxy error.
Features
- Added public Pinterest profile URLs as an input mode, including profile metadata and profile pin extraction.
- Added source provenance and automatic deduplication when searches, profiles, boards, or pin URLs overlap.
Reliability
- Added a reliable default proxy route for cloud runs when Pinterest returns empty responses from datacenter IPs.
- Empty extraction results now fail clearly instead of reporting a misleading successful run.
Features
- Added public Store task examples so users can start common Pinterest scraping workflows faster.
Changes
- Made the default health-check input lighter so sample and monitoring runs finish with less unnecessary scraping.
Reliability
- Improved reliability for Pinterest search, board, and pin runs with hardened retry handling.
- Restored per-pin charging for successfully scraped pin records so pricing matches dataset output.
Features
- Initial release with support for scraping Pinterest search results, boards, profiles, and pin pages, including pin metadata, images, pinner and board details, hashtags, source domains, and outbound links.
