An AliExpress scraper API turns a product keyword into structured listings — product id, title, sale price and original price, discount percent, star rating, orders sold, store name and link, ships-from, selling points, image and item URL.
The field that makes AliExpress worth scraping is orders. Almost no other marketplace publishes a units-sold figure next to every listing, and it is the closest thing to public demand data that exists at this scale — which is why product researchers, dropshippers and trend analysts keep coming back to it. This collector delivers it as a number, alongside the sale-versus-original price pair that tells you whether a discount is real.
$0.001 per delivered item, up to 250 items per run. Nothing delivered means nothing charged.
AliExpress renders its result cards from a JSON hydration payload embedded in the page rather than from server-side HTML. That is good news — the data is already structured before anyone parses it — and it is also where most home-grown AliExpress scrapers break, because they hard-code the path to the item node and AliExpress reshuffles that structure regularly.
This collector locates the item node by shape instead of by path: it looks for the object that has the fields an item has, wherever it currently lives. A reorganisation of the payload therefore does not break the run. No browser is involved at any point, which is why a 250-item run costs a quarter.
query is required. currency is more consequential here than on other marketplaces: AliExpress prices are quoted in the visitor's currency, so a series collected without pinning it will drift with exchange rates and look like price movement that never happened. country sets the exit and the regional site together.
Each item is a row. The pair to watch is price_value against original_price: AliExpress discounts heavily and permanently, so discount_percent is better read as a positioning signal than as a sale. orders is the cumulative units-sold figure the listing displays — the single most useful field on the page for demand research. store and store_url let you pivot from an item to everything that seller lists, and ships_from is what separates a two-week delivery from a two-day one.
$0.001 per delivered item ($1 per 1,000). Nothing delivered means nothing charged, and the $2 monthly free credit covers roughly 2,000 items here. Volume tiers take up to 30% off.
$0.001 per delivered item, $1 per 1,000. Product research is usually wide and shallow — 300 keywords at 60 items each is 18,000 rows, $18 at list price — and because you pay per delivered item rather than per request, a keyword that returns twelve results costs twelve tenths of a cent, not a full page fetch.
Blocked pages and retries are never billed.
AliExpress publishes an affiliate/open platform API, and like Amazon's it is built for a commercial relationship rather than for research. You register an application, get approved, and work inside a product surface designed to help you promote items and earn commission. Its coverage is the affiliate catalogue, not the search page, and its ranking is not the ranking a shopper sees.
For demand research the practical gap is the orders figure and the live sale price in the position AliExpress actually served them. That is a property of the result page. If you want to know what a buyer searching "running shoes" sees today, the page is the answer and the affiliate catalogue is a different dataset that happens to overlap.
The public AliExpress scrapers on GitHub have a characteristic failure mode: they work beautifully for a few weeks and then return empty arrays. Almost always the cause is the hydration payload moving, because the scraper walked a fixed path into it.
If you would rather own the fetcher, the residential exits underneath this collector are available on their own.
Sort a category's results by orders to see what is actually selling rather than what is merely listed. Re-run weekly and the delta in orders is a demand curve.
Filter for high orders, decent rating and a ships_from that matches your delivery promise. The store link gives you the rest of that supplier's catalogue in one hop.
Run the same product keyword here and on the eBay or Amazon collectors to see the spread between source and retail.
Watch your brand terms for listings you did not authorise; store and selling_points usually identify the pattern quickly.
Up to 250 items per run, pagination handled internally, the default is lower — set max_results to what you will actually use, since you pay per delivered item. Throughput follows your plan's rate limit rather than the collector: 60 requests/minute on pay-as-you-go, up to 1,200 on the top tier.
Legally: listing data is public and collecting public data is generally lawful in most jurisdictions, but AliExpress's terms restrict automated access and the position varies by country. Store names can be personal data when the seller is an individual. This is not legal advice.
There is a free monthly allowance: $2 of usage with no card, about 2,000 delivered items at $0.001 each. That is enough for a full round of product research before you spend anything. The free scrapers on GitHub are real but tend to break when AliExpress reshuffles its hydration payload.
Yes — orders is the cumulative units-sold figure AliExpress displays on the listing, delivered as a number. It is the reason most people scrape AliExpress at all, since almost no other marketplace publishes anything comparable.
$0.001 per delivered item, $1 per 1,000, with no subscription. Volume tiers take up to 30% off and GET /v1/scraper/collectors returns the price your key actually pays. Blocked pages and retries are never billed.
Usually because currency was not pinned. AliExpress quotes prices in the visitor's currency, so a run without an explicit currency will drift with exchange rates and look like price movement. Set currency explicitly for any series you intend to compare over time.
No. The listing data comes from a JSON payload embedded in the page, so the run is a plain request and a parse. That is why it costs a tenth of a cent per item instead of a cent.
This collector is keyword-driven, but every row carries store and store_url, so a search gives you the seller identity you need to pivot. Searching brand or store terms is usually the fastest route to a supplier's visible catalogue.
The affiliate API serves the affiliate catalogue for promotion, after an application approval, with its own ranking. This returns the public search page as served in a country, including the orders and discount figures in the position AliExpress actually showed them.
Skip the marketing. QuantumProxies.io publishes a machine-readable site map, Markdown for every page, a free MCP server, and APIs billed from the same balance as your proxies.
Connect in one command: npx -y quantumproxies-mcp