Taobao & Tmall Scraper - Products, Prices, Sellers & Reviews

Scrape Taobao and Tmall - China's largest consumer marketplaces - by keyword, category, product ID or shop. Get prices in 13 currencies, full seller profiles with DSR service scores, buyer-rating breakdowns, real customer reviews and complete category paths.

No login. No cookies. No browser. Pure HTTP against Taobao's own server-rendered catalogue, so runs are fast, cheap and don't depend on a headless browser that can break overnight.

---

Why this actor exists

Taobao is the hardest large marketplace on the web to scrape. Its consumer pages sit behind Alibaba's x5sec / baxia wall: s.taobao.com/search serves a client-only shell with zero product data, and item.taobao.com, detail.tmall.com and list.tmall.com answer with a punish interstitial. That is why the leading Taobao actor on Apify runs at roughly two-thirds success - buyers are already paying for something that fails one run in three.

This actor doesn't fight that wall. Taobao publishes the same catalogue, fully server-rendered, for search engines on world.taobao.com and its country hosts, and its own robots.txt explicitly allows crawlers on /product/, /category/, /item/ and /dianpu/. Reading that surface measured 100% success across 24 consecutive requests on four different lanes.

---

What you get

Products (search, category, item modes)

Group	Fields
Identity	itemId, title, url, taobaoUrl, platform (taobao / tmall), itemIdStr
Pricing	price + currency in your region's money, priceCny, priceHkd, priceTwd, originalPrice, promotionPrice, discountPrice, discountType, hasRangePrice, priceNote
Seller	shopId, shopName, sellerId, sellerType + sellerTypeLabel (individual vs business), sellerCreditLevel, shopOpenedAt, shopItemCount, shopNewItemCount, shopUrl, shopIcon
Seller quality (DSR)	dsrDescription, dsrService, dsrDelivery - Taobao's three published 1-5 seller scores
Ratings & reviews	reviews[] (rating, text, date, variant bought, reviewer), reviewTags[] (aggregated sentiment tags with counts), and — on Taobao listings only, see the note below — ratingGoodPercent, ratingBadPercent, ratingNeutralPercent, ratingGoodCount, ratingBadCount, reviewCount
Classification	categoryId, rootCategoryId, categoryName, categoryPath, breadcrumbs[], descriptionMatchScore
Logistics	location (ships-from), freeShipping, areaLimit, paymentMethods[]
Media	mainImage, images[], imageCount

⚠️ Read this before you buy: the rating breakdown is Taobao-only

Tmall listings do not publish the buyer-rating breakdown. Taobao does. This is Tmall's own decision about what it exposes publicly - not a gap in this scraper - and no tool can return those numbers for a Tmall listing without a logged-in session.

Measured on a real 30-product run: the ratingGoodPercent / ratingBadPercent / ratingNeutralPercent / ratingGoodCount / ratingBadCount / reviewCount fields came back on 12 of 12 Taobao listings and 0 of 10 Tmall listings. Roughly a third to a half of a mixed keyword search is typically Tmall, and more in brand-heavy categories, so expect those six fields to be empty on a meaningful share of rows.

What you still get on every Tmall row (same run, 10 of 10): a sample customer review with its star rating, all three seller DSR scores (dsrDescription, dsrService, dsrDelivery) and descriptionMatchScore. So you can still rank and filter Tmall sellers on quality - just not on a per-item positive-review percentage.

Use the platform field (taobao / tmall) to tell the two apart in your own filters.

Shops (shop mode)

shopId, shopName, sellerId, logo, all three DSR scores plus dsrAverage, the shop's listed products with prices, categoryPath, and a 0-100 leadScore - a ready-made seller-prospecting list.

---

Modes

Mode	Give it	Use it for
Search by keyword	keywords	Product research, price benchmarking, finding suppliers
Browse a category	categoryIds	Full-category sweeps; sub-categories are followed automatically
Specific products	itemUrls (IDs or any Taobao/Tmall URL)	Price monitoring on a watchlist
Shops / seller leads	shopIds	Building a seller list for sourcing or outreach

Paste any Taobao, Tmall or world.taobao.com URL into Product IDs / URLs and it is routed to the right mode automatically.

---

Regions and currency

Pick the storefront you sell or buy in and prices come back in that currency, with the yuan price always alongside in priceCny:

Singapore (SGD) · Global/HK (CNY) · Malaysia (MYR) · Thailand (THB) · Vietnam (VND) · Philippines (PHP) · Cambodia (KHR) · Macau (MOP) · Australia (AUD) · New Zealand (NZD) · South Korea (KRW) · Japan (JPY) · Canada (CAD)

---

Example input

{
  "mode": "search",
  "keywords": ["手机", "connector"],
  "region": "sg",
  "maxItems": 200,
  "scrapeDetails": true,
  "deepScan": true
}

Chinese keywords return the most results - Taobao is a Chinese-language catalogue - but English terms work too.

Example output (one product)

{
  "type": "product",
  "itemId": "1010692564493",
  "title": "新款老人机2026年5G全网通4G老年人手机大字大声超长待机备用学生",
  "url": "https://world.taobao.com/item/1010692564493.htm",
  "platform": "taobao",
  "price": 35.8,
  "currency": "SGD",
  "priceCny": 185,
  "priceHkd": 216.47,
  "discountType": "百億補貼",
  "shopName": "十三手機通訊",
  "sellerTypeLabel": "Individual seller",
  "sellerCreditLevel": 18,
  "dsrDescription": 4.8,
  "dsrService": 4.9,
  "dsrDelivery": 4.9,
  "ratingGoodPercent": 99,
  "reviews": [
    { "rating": 5, "content": "寶貝收到，快遞迅速，品質不錯，下次再買。", "date": "2026.02.07 08:53", "sku": "顏色:黑色", "user": "***" }
  ],
  "categoryPath": "首頁 > 手機通訊 > 手機 > 老人手機",
  "location": "廣東深圳",
  "freeShipping": true
}

---

How deep can it go?

Taobao's crawlable surface serves one page of ~40 results per query - there is no deep pagination on it. Rather than pretend otherwise, Deep scan re-queries each keyword across sort orders and price bands to assemble a much larger unique set: measured 279 unique products from a single keyword versus 40 for one page. Turn on Follow related keywords to widen further via the related searches Taobao links from each page, or use category mode, which walks the sub-category tree.

Honest limits

• Tmall listings do not publish the buyer-rating breakdown - measured 12/12 on Taobao rows vs 0/10 on Tmall rows. Full explanation and what you still get on Tmall rows is in the callout above. Filter on platform to separate them.
• One primary image and one sample review per listing. The full gallery and full review pagination are only behind Taobao's browser-only API. imageCount reflects what is genuinely there.
• Review counts are approximate where Taobao itself publishes them rounded ("10,000+").
• Use the RESIDENTIAL proxy. Taobao rate-limits single IPs hard; datacenter measured 33-67% on the keyword lane against 100% on residential. Every retry automatically rotates to a new IP.

---

Use with AI assistants (MCP)

This actor works as a tool for AI assistants and agents through the Apify MCP server - ask Claude, ChatGPT or your own agent to pull Taobao pricing, seller quality or category data and it can call this scraper directly.

Agent-ready: autonomous payments (x402 & Skyfire)

This actor is agent-ready — AI agents can discover it, run it, and pay for it autonomously, with no Apify account and no human in the loop. It uses pay-per-event pricing and limited permissions, so it qualifies for Apify's agentic-payment standards:

• x402 — an open, HTTP-native payment protocol. Agents pay per run in USDC on the Base network directly through the Apify MCP server — no account, no API key.
• Skyfire — agent-to-service payments for fully autonomous AI-agent workflows.

Building an AI agent, MCP tool, or autonomous data pipeline? This scraper is ready to plug in and pay as it goes.

---

Pricing

Pay per result, no monthly platform fee and no start fee:

Event	Price
Product scraped (listing fields)	$0.003
Product with full detail (seller profile, ratings, reviews, breadcrumbs, location)	$0.006
Shop / seller lead	$0.008

A product is charged either as a listing or as a full-detail record - never both.

Legal

Scrapes only publicly available pages that Taobao itself serves to search-engine crawlers, on the exact paths its robots.txt allows them. No login, no personal data behind an account wall. You are responsible for how you use the data, including compliance with applicable terms and data-protection law.