Every AliExpress price.
In the market you ship to.

No. There is no affiliate program to join, no open-platform application to get approved, no app key or request signing, and no AliExpress sign-in. Your FetchLayer key is the only credential, and it reads what AliExpress publishes to any shopper: search results, category listings, product pages, prices, stock and buyer reviews. FetchLayer is not affiliated with, endorsed by, or sponsored by AliExpress or Alibaba Group.

The ones you ask for. Every scraping endpoint takes shipTo (a two-letter country code), currency and language, and every result echoes the market it was read in as market. Unset, they default to US, USD and en_US. The accepted currencies are USD, EUR, GBP, CAD, AUD, NZD, BRL, MXN, CLP, COP, PEN, ARS, JPY and KRW, and the languages are en_US, de_DE and pt_BR — values verified to work one at a time. Anything else is a 400, because AliExpress itself answers an unsupported currency with an empty page that looks exactly like "no matches".

Because AliExpress changes the content, not just the formatting. The same keyword ranks different products first for a US, German and British shopper, the same product is priced in a different currency, and titles are translated. So when you monitor prices, pin shipTo and currency and keep them fixed: a price read in a different market is a different price, not a price change.

Up to 20 product ids or URLs per request to /product-prices. Each comes back in request order with its own status — ok, unavailable, not_found, or failed, meaning it could not be checked this time — plus the lowest and highest variant price, original price, total stock, and one variant’s price and stock. Name a variant with a skuId to track it. One product that cannot be checked never fails the rest, and the whole batch is one request against your plan.

Product details and price checks depend on product lookups that AliExpress refuses in bursts. When that happens the request returns a 503, which is temporary: wait a little and retry, and it usually goes through. Search, category listings and reviews are served another way and are not affected. Do not retry a 400, which names the field to fix, or a 404, which means the product or category does not exist.

Yes. /product-reviews returns 20 reviews a page with the rating summary — the average, the total, the positive share and the count at each star — and how many reviews have photos or a follow-up. Each review has its rating, date, text translated into your language where AliExpress did so and the reviewer’s original words, masked reviewer name and country, the variant bought, photos, any follow-up and the seller’s reply. Many AliExpress reviews are a rating with no text; those come back with text null.

As complete as AliExpress lists them. It shows at most 60 pages for one query, so narrow with price, shipping, rating, Choice and ship-from filters to reach further. Consecutive pages can repeat a product, so de-duplicate by productId across responses, and a search with few matches is padded by AliExpress with loosely related products — the notes array says so when it happens.

Yes. AliExpress numbers products differently on its US storefront, and those ids are accepted everywhere a product is. Every id the API returns is the global one, so the same product always comes back under the same id whatever market it was read in — which is what makes a price history keyed on productId safe.

One API request equals one credit, the same as every other FetchLayer platform. On /product-prices each product priced counts as one request, so 20 products bill as 20; a product that fails on our side is not billed. On search, category listings and reviews each page walked counts as one request, so pages: 5 bills as five. Parsing a URL with /resolve-url makes no request to AliExpress and is free. Streaming an image or video bills as a normal request. AliExpress is in Beta while coverage expands, billed at half credit.