Skip to main content
​
June 16, 2026

​
Improved

  • Google Maps Leads: new filters and result fields — added max_rating (upper bound on star rating), min_reviews, max_reviews (review count range), exact_name_match (target a specific brand or chain), phone_filter, and verified_filter (owner-claimed listings). New result fields: owner_posts, customer_posts, image_categories, menu_items, is_verified, kgmid, reviews_link, name_for_emails, streetview_url, county, state_code, region, timezone, owner_name, owner_link, popular_times, popular_times_live, booking_link, actions, order_providers, reviews_tags, and people_also_search.
  • Twitter Search Results: fetch_since support — use a relative duration (24h, 7d, 2w) or an absolute date (YYYY-MM-DD HH:MM:SS) to stop scraping when tweets older than the threshold are reached. Works best with Latest search URLs (&f=live). Also added fetch_since_timezone and max_unique_results_per_run.
  • LinkedIn Posts Commenters & Likers: max_results and max_unique_results_per_run — cap the number of engagers collected per post or across the entire run.

​
Fixed

  • Sales Navigator Profile: positions scraping — work experience entries were being parsed incorrectly. Fixed.
  • Sales Navigator Companies: re-login crash — an AssertionError after automatic re-login no longer interrupts runs.
  • Sales Navigator email enrichment — enrichment requests were failing silently in some cases. Fixed.
  • LinkedIn Posts Commenters & Likers: result deduplication — commenters/likers were occasionally duplicated in exported results. Fixed.
  • TikTok Top Ads: industry filter IDs — incorrect internal IDs were being sent for several industry categories, causing filter to have no effect. Fixed.
  • Leboncoin listings: 0 results — a parsing change on Leboncoin caused some search pages to return 0 results. Fixed.
  • Leboncoin Boutiques: escape sequence parsing — invalid escape sequences in store descriptions caused parsing errors. Fixed.
  • ParuVendu listings: pycurl stream error — intermittent stream errors on listing pages are now caught and retried. Fixed.
  • PAP: connection retries — base-level ConnectionError exceptions from curl_cffi are now retried automatically. Fixed.
  • LaCentrale Listing Status Checker: DataDome bypass — migrated to a more reliable DataDome bypass path; improves success rate on listing status checks.

​
Platform

  • Export: JSON and JSONL formats — run results can now be downloaded as .json (array) or .jsonl (newline-delimited JSON) in addition to CSV and XLSX.
  • Export: field selection — use fields to select a subset of columns, or omit to exclude specific columns from exported files.
  • Export: XLSX download — XLSX is now available as a direct download format when exporting run results.
  • Export: faster XLSX generation — switched to xlsxwriter for XLSX exports; large files generate significantly faster.
  • Export: background conversion for large files — XLSX exports over Excel’s row limit (~1 M rows) are rejected upfront with a clear error. Large exports are offloaded to a dedicated background queue so the API response is not blocked.
  • Export: live progress — background export conversion progress is now surfaced in real time.