1688 Wholesale Scraper API

Search 1688.com — Alibaba's domestic Chinese wholesale marketplace — and get structured product rows back: offer ID, title, unit price, the full quantity price ladder, minimum order quantity, the supplier company, sales counts, the seller's service rating, product image and product URL.

No 1688 account, no cookies, no captcha solving, no proxy of your own. Type a search term, get rows.

Price

$0.35 per 1,000 products, plus $0.001 to start a run.

That is the lowest price for 1688 product data on Apify. The cheapest other 1688 scraper charges $0.45 per 1,000; the cheapest one with a real user base charges $0.63; the most-installed one charges $4.99. You are charged per product row returned — a run that finds nothing costs only the start fee, and diagnostic rows are never charged.

Products	Cost
500	$0.18
2,000	$0.70
10,000	$3.50
100,000	$35.00

What a row looks like

A real row from a run on 手机壳 (phone case):

{
  "offerId": "822216254820",
  "title": "适用苹果16PROMAX皮纹简约纯色手机壳iphone11/13防摔tpu保护套",
  "productUrl": "https://detail.1688.com/offer/822216254820.html",
  "price": 1.5,
  "priceMin": 1,
  "priceMax": 1.5,
  "currency": "CNY",
  "priceTiers": [
    { "minQuantity": 1,     "maxQuantity": 499,   "quantityLabel": "1~499",     "price": 1.5 },
    { "minQuantity": 500,   "maxQuantity": 99998, "quantityLabel": "500~99998", "price": 1.2 },
    { "minQuantity": 99999, "maxQuantity": null,  "quantityLabel": "≥99999",    "price": 1 }
  ],
  "minOrderQuantity": 1,
  "unit": "个",
  "salesCount": 14821,
  "repeatPurchaseRate": "13%",
  "rating": 4.5,
  "ratingBreakdown": { "goods": 3.67, "logistics": 3.57, "consultation": 4.5, "dispute": 4, "afterSales": 4 },
  "supplierName": "广州市艾尚数码科技有限公司",
  "supplierUrl": "https://m.1688.com/winport/b2b-2206457435068a76ea.html",
  "supplierProvince": "广东",
  "supplierCity": "广州市",
  "supplierYears": 7,
  "supplierId": "b2b-2206457435068a76ea",
  "businessType": "生产加工",
  "isFactory": true,
  "buyerProtections": ["7天包换", "48小时发货", "诚e赊"],
  "imageUrl": "https://cbu01.alicdn.com/img/ibank/O1CN01NThc6r1nJCEvpDbJH_!!2206457435068-0-cib.jpg",
  "isAd": false,
  "searchKeyword": "手机壳",
  "page": 1,
  "position": 2,
  "scrapedAt": "2026-08-09T06:33:01.546Z"
}

The price ladder is the reason most people scrape 1688: price is what one unit costs at the smallest order, and priceTiers is the discount schedule as the order grows. priceMin and priceMax are the ends of that ladder, so you can sort by best-case unit cost without unpacking it.

Field coverage

Measured over a real 500-row run, not estimated:

Field	Present
offerId, title, price, productUrl, minOrderQuantity, unit	100%
salesCount, repeatPurchaseRate, rating, ratingBreakdown	100%
supplierName, supplierId, supplierYears, imageUrl	100%
buyerProtections	99%
supplierProvince, supplierCity	98%
businessType	85%
priceTiers	74%

priceTiers is empty on the rest because those sellers list a single price with no quantity breaks — an empty array is the honest answer, so no tier is invented. Missing values are null, never a guess.

Input

Field	What it does
searchTerms	Search terms. Chinese terms match far more than English ones — 手机壳 returns a fuller catalogue than "phone case", though both work.
startUrls	Or paste 1688 search URLs. 1688 encodes its URLs as GBK rather than UTF-8, which this decodes correctly.
maxItems	Total products across all terms. Default 40.
sortBy	relevance (default), priceAsc, priceDesc.
minPrice / maxPrice	Price bounds in CNY.
requestDelayMs	Gap between requests. Default 1500; lower it to go faster.

{
  "searchTerms": ["手机壳", "连衣裙"],
  "maxItems": 2000,
  "sortBy": "priceAsc",
  "maxPrice": 50
}

Run it with no input and you get one labelled sample row, uncharged, so you can see the shape before spending anything.

Limits

Read this part before you build on it.

• 2,000 products per search term. 1688 stops serving results at that depth no matter how you page through them. It reports "2,000 matches" for any broad term, which is a ceiling, not a count. To go wider, pass more terms — narrower ones return different products, not the same 2,000.
• sortBy: "booked" is not offered. 1688 accepts a best-selling sort and then returns salesCount: 0 on every row it gives back — 100 of 100, against 96 of 100 on relevance. Shipping it would mean handing you a sales ranking whose sales figures are all zeroes.
• Prices are CNY, ex-factory. No shipping, no tax, no export handling. A 1688 price is what a domestic Chinese buyer pays at the warehouse door.
• About 1–2% of rows are paid placements. They are real listings and are marked isAd: true, so you can filter them rather than wonder about them.
• No per-product detail fetch. Everything above comes out of the search response, which is why a 2,000-row pull is 40 requests instead of 2,040. Full specification tables and description images live on the product page and are not included.
• Titles and supplier names are Chinese. They are returned verbatim, untranslated.
• minOrderQuantity: 0 means the seller stated no minimum, not that the field is missing.

FAQ

Do I need a 1688 or Alibaba account? No. Nothing to log in to, and nothing to solve.

Do I need Chinese proxies? No — and this is the part most 1688 scrapers get wrong. See "How it works" below.

How fast is it? 1,941 products in 71 seconds in testing, on the default settings.

What does it cost to scrape 1,000 products? $0.35, plus the $0.001 start fee.

Can I search in English? Yes, and it returns real results. Chinese terms return substantially more of the catalogue, because that is how sellers write their titles.

Is 1688 the same as Alibaba.com? No. Alibaba.com is the English-language export site; 1688 is the domestic Chinese one, generally with lower prices and lower minimum orders, aimed at buyers inside China.

Can I export to CSV or Excel? Yes — Apify exports the dataset as JSON, CSV, Excel or XML, with a REST endpoint for the same data.

Is this affiliated with 1688 or Alibaba? No. It reads public listing data. Check 1688's terms and your own local rules before using the output commercially.

How it works

1688's website sits behind Alibaba's "punish" gate, and it is unusually aggressive. A US datacenter address is challenged on the first request. Renting mainland-Chinese addresses does not fix it either: five Guangdong addresses tested against the search page ran clean for two or three requests each and then returned a challenge page on everything afterwards, still blocked half an hour later. This is why 1688 scrapers are typically expensive — they pay for headless browsers and residential Chinese bandwidth to fight that gate, and pass the cost on.

This scraper does not touch the website. It talks to the JSON API behind 1688's mobile site, which is a different gate entirely — the same addresses that were permanently blocked on the search page answer it normally, and an ordinary datacenter address runs forty-plus pages without a challenge. No browser is launched, no Chinese IP is rented, and no residential bandwidth is billed.

That is the whole reason this can be the cheapest 1688 scraper on the platform and still be profitable: the expensive part of the problem is avoided rather than paid for.