Taobao & Tmall Product Scraper: prices, sellers, reviews, no login

Extract products from Taobao (淘宝) and Tmall (天猫) by keyword, category, shop or item id, and get clean JSON, CSV or Excel: title, promo and original price in CNY converted to USD, EUR, GBP, SGD, MYR, AUD, JPY, KRW, THB, CAD, HKD and TWD, shop, category, seller type and credit level, the seller's three DSR scores, ships-from city, review tag counts and a sample review.

No Taobao account, no cookies, no browser. Datacenter proxies are enough. English field names throughout, so you don't need to read Chinese to use the columns.

This Actor is an independent tool and is not affiliated with, endorsed by or sponsored by Alibaba Group. Taobao and Tmall are referenced only to describe the public website it reads.

What it does

Give it	Get
Keywords (手机壳, 奶粉, iphone case)	The products Taobao lists for the keyword, plus the products of related keywords buyers actually search (deep scan, see below)
Category ids (1512 = phones)	The category's listed products, plus related keyword pages
Shop ids (539056969)	The shop's catalog, up to 60 products, each with the shop's seller scores
Item ids or product URLs	One row per item with live price, seller details and a sample review
Open each product page (option)	Adds seller type, credit level, DSR scores, location, category path, review tags, payment methods and original vs promotion price to every row

Prices are the real prices shown to shoppers, not the masked ones on the product page.

Why this scraper

• It reads the pages Taobao publishes for search engines. Taobao's consumer pages sit behind a login wall and Alibaba's bot defenses; its SEO catalog on world.taobao.com is server-rendered, carries the product data as JSON, and is explicitly allowed for crawlers in Taobao's own robots.txt. Reading that surface is why runs are fast and dependable.
• Retries rotate IPs automatically. Taobao throttles single IPs hard. Every throttled response is retried on a fresh IP without any configuration. On our test runs, 181 of 181 requests succeeded.
• Deep scan finds more than the 20 products per page. Taobao's crawlable catalog ignores paging, sorting and filters: one keyword page is one page of 20. Instead of pretending otherwise, deep scan asks Taobao's own search-box autocomplete for the related keywords buyers use (奶粉 → 奶粉3段, 奶粉试用装 ...), reads each of those keyword pages too, follows the breadcrumb category, and dedupes. On our run of 奶粉 that turned 20 products into 60 in under a minute; 手机壳 reached 108 with 15 extra pages. Every row says which keyword it was found through (foundVia).
• One row per product, with English column names. Chinese text values are kept as Taobao shows them; everything else is typed and named for a spreadsheet.
• Cheap to run. Pay per product, no platform usage charged on top. A 150-product run with every product page opened cost about two cents of platform usage on our tests.

Pricing

Pay per event. You are not charged for platform usage.

Event	Price
Product returned	$2.50 per 1,000
Product page opened (only with Open each product page on)	+$2.50 per 1,000

So a product list costs $2.50 per 1,000 rows, and a fully detailed list $5 per 1,000 rows.

Input

{
    "keywords": ["奶粉", "iphone case"],
    "shopIds": ["539056969"],
    "itemIds": ["1083028793486"],
    "maxItemsPerQuery": 100,
    "deepScan": true,
    "includeDetails": true
}

Field	Type	Default	What it does
keywords	array of strings		Search words, Chinese or English. Chinese keywords give the widest results and the best deep scan.
startUrls	array of {url}		item.taobao.com/item.htm?id=..., detail.tmall.com/item.htm?id=..., world.taobao.com/item/..., world.taobao.com/product/<keyword>.htm, world.taobao.com/category/<id>.htm, world.taobao.com/dianpu/<shopId>.htm, s.taobao.com/search?q=...
itemIds	array of strings		Item ids, one row each
categoryIds	array of strings		Taobao category ids (categoryId column of results, or world.taobao.com/category/<id>.htm)
shopIds	array of strings		Shop ids (shopId column of results, or world.taobao.com/dianpu/<id>.htm)
maxItemsPerQuery	integer	100	Unique products per keyword, category or shop
deepScan	boolean	true	Read related keyword pages and the breadcrumb category to get past 20 products
maxFacetPages	integer	40	Upper bound on extra pages per query
includeDetails	boolean	false	Open each product page for seller and review fields
region	select	Global	Overseas storefront to read (Singapore, Malaysia, Japan ...). Prices stay in CNY with HKD/TWD alongside.
maxConcurrency	integer	10	Parallel requests
proxyConfiguration	proxy	Apify datacenter	Residential is not needed for this site

Output

One row per product. Listing fields come on every row; detail fields come when Open each product page is on or when you pass item ids.

Identity: itemId, itemIdStr, url, taobaoUrl, title

Price: priceCny (promo price), originalPriceCny, priceHkd, priceTwd (Taobao's own conversions), plus priceUsd, priceEur, priceGbp, priceSgd, priceMyr, priceAud, priceJpy, priceKrw, priceThb, priceCad (converted at the day's rate, fetched at run start), priceLocal, priceCurrency, discountType, freeShipping, areaLimit

Shop and category: shopId, shopName, categoryId, rootCategoryId, descriptionMatchScore, reviewTags

Media: imageUrl, images

Context: queryType (keyword / category / shop / item), query, foundVia (the related keyword or category page the product came from), position, totalResultsForQuery (how many products Taobao reports for the query), scrapedAt

Shop mode adds: sellerId, shopDsrDelivery, shopDsrService, shopDsrDescription

Detail fields: detailStatus, platform (taobao / tmall), location (ships from), categoryPath, promotionPriceCny, hasRangePrice, priceNote, detailImages, seller {shopId, shopName, sellerId, sellerType (business / individual), creditLevel, itemCount, newItemCount, shopOpenedAt, shopIcon, dsrDelivery, dsrService, dsrDescription}, sampleReview {rating, text, date, sku, user}, hasMoreReviews, reviewCountText, reviewSummary [{count, tag}], paymentMethods, sellableLanguages

Example row (detail on):

{
    "itemId": "1083028793486",
    "title": "【需當面激活丨12期免息】Apple/蘋果 iPhone 18 Pro Max 手機 2026新款 全新國行 官方正品全國聯保",
    "url": "https://world.taobao.com/item/UmhOMm04UFJRL2ZYY0FmdnIxdnp6Zz09.htm",
    "taobaoUrl": "https://item.taobao.com/item.htm?id=1083028793486",
    "priceCny": 10999,
    "priceUsd": 1545.26,
    "priceEur": 1329.88,
    "priceHkd": 13075.61,
    "priceTwd": 53984.19,
    "discountType": "優惠價",
    "shopId": "539056969",
    "shopName": "Apple產品健康專營店",
    "categoryId": "1512",
    "platform": "tmall",
    "location": "江蘇南京",
    "seller": { "sellerType": "business", "creditLevel": 8, "itemCount": 54, "dsrDelivery": 4.9, "dsrService": 4.8, "dsrDescription": 4.9 },
    "sampleReview": { "rating": 5, "text": "發貨速度快，隔天就到了。而且國補後 很優惠。", "date": "2026.09.20 10:33" },
    "reviewSummary": [{ "count": 21, "tag": "物流速度很快" }, { "count": 22, "tag": "性價比高" }],
    "queryType": "keyword",
    "query": "iphone"
}

Tmall only?

Want just Tmall's official brand stores? Turn on Open each product page and every row gets a platform field set to tmall or taobao. Filter on it in the dataset view, Excel or your code. You can also pass a Tmall flagship store's shopId to get that store's catalog directly.

Honest limits

• 20 products per listing page. Taobao's crawlable catalog ignores page, sort and filter parameters. Deep scan works around this with related keyword pages; the number of unique products per keyword depends on how many related keywords Taobao suggests (Chinese keywords get the most).
• Titles on item-id rows can be shortened (Taobao truncates the title on the detail page). Rows found through a keyword, category or shop carry the full title.
• One sample review per product and aggregated review tag counts. Full review pagination and the full image gallery are only behind Taobao's browser-only API.
• Tmall listings don't publish the buyer rating breakdown, so reviewSummary is empty on some Tmall rows. You still get the seller DSR scores and a sample review.
• English keywords work when Taobao has a page for them (iphone case does); their autocomplete suggestions often don't, so deep scan adds less. Use Chinese keywords for breadth.

Use with AI agents (MCP)

This Actor works as a tool through the Apify MCP server. Ask Claude, ChatGPT or your own agent to pull Taobao prices, seller quality or a shop's catalog and it can call this scraper directly. It uses pay-per-event pricing with limited permissions, so it also qualifies for agentic payments (no Apify account needed).

Use via API

from apify_client import ApifyClient

client = ApifyClient("<YOUR_API_TOKEN>")
run = client.actor("<ACTOR_ID>").call(run_input={
    "keywords": ["奶粉"],
    "maxItemsPerQuery": 200,
    "includeDetails": True,
})
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item["title"], item["priceCny"], item["shopName"])

Legal

This scraper reads publicly available product listings that Taobao publishes for search engines, in line with Taobao's robots.txt. It does not log in, bypass CAPTCHAs or access private data. You are responsible for how you use the data; respect Taobao's terms and applicable law, and don't collect personal data.