- Fixed: a short run gave up after one proxy attempt. The time guard priced
every retry at its worst case — a full 60-second timeout — and refused to
start a retry it could not fit five times over. On a run with about a hundred
seconds left that allowed exactly one attempt, so the run asked once, met
Alibaba's slider challenge in a fraction of a second, and stopped there
without ever rotating to the fresh exit IP that clears it. The actor now
measures the time each attempt actually takes: every exit and browser
fingerprint is tried while the clock allows, and only a genuinely spent run
stops early.
- Fixed: a run that ran out of time said it had been challenged. The status
message blamed Alibaba's velocity gate "after every proxy exit and browser
fingerprint was rotated" when a single attempt had been made. A run stopped by
the clock now reports the clock — "N of M proxy attempts tried before the
crawl box closed" — says plainly that it cannot tell whether Alibaba would
have served the page, and charges nothing for it. The gate verdict is reserved
for a rotation that really was walked to the end.
- The run now watches its own clock. Long searches used to be killed by the
platform timeout mid-page, losing every product that had not been pushed yet.
The actor now stops walking pages while there is still time to charge and
deliver what it collected, and ends successfully with a status message
saying "Stopped at the run time limit: N rows delivered, raise the run
timeout or lower maxResults."
- Page and supplier-storefront fetches are bounded by the time actually left,
not by a fixed retry ladder. A single slow page could previously burn five
minutes of retries past the point where the run still had time for it.
- Honest limits in the input form: Max products now tops out at 4,000
(100 result pages × ~40 products — the real depth of one Alibaba search)
instead of advertising 100,000 that no run could reach, and Search terms or
URLs accepts up to 20 entries per run. With Add supplier trust profile
on, the realistic reach of an hour-long run (~800 products) is now stated.
- Every run now ends with a status message instead of finishing silently — and
a run that was challenged, cut short by the clock or stopped by your charge
limit says which, instead of looking like an empty search.
- README: added the missing Pricing section, disclosing per-event prices and the $0.001 actor-start fee.
- Invalid input no longer fails the run. The run now finishes cleanly with a clear explanation in its status message, so a typo or an empty list is easy to spot and fix.
- List inputs now accept datasets piped straight from another scraper: entries
that are objects are read through their usual key names (url/link,
query/keyword/title, name/company and similar) instead of being rejected.
- Moved to one flat all-in price: platform usage is now included, so there is a
single price per product delivered and no separate usage bill to reconcile.
- Removed the proxy setting from the input. The actor now handles finishing a run
entirely on its own, with nothing to configure.
First release. Scrape Alibaba.com by keyword, search URL or product URL and get:
- Product title, link, image, price and price range, currency, and tiered MOQ
pricing when the listing publishes a price ladder.
- Minimum order quantity, both as a number and as shown.
- Supplier name, storefront link, country, and Verified membership tenure.
- Trust signals: Verified Supplier, Trade Assurance, top/crown supplier, and the
product certifications shown on the listing (CE, FCC, RoHS and the like).
- Buyer review score and count.
- Optional supplier trust profile: opens each supplier's storefront for Verified
tenure, Trade Assurance, response rate and business type. Billed only when a
profile is returned.
Paced and rotating so a search finishes with a full set of results instead of
stalling partway. Flat JSON output, no API key.
