Global Marketplace & Variant Specialist

Extract every product variant — size, color, material — from the world's most anti-bot-protected marketplaces.

Built for dropshippers, competitor analysts, and feed managers who need clean, structured SKU data from fast-fashion and global marketplace sites — not raw HTML to parse yourself.

---

Supported sites

Site	Extraction method	Anti-bot level
AliExpress	window.runParams (full SKU matrix)	★★★★
Temu	__NEXT_DATA__ hydration JSON	★★★★★
SHEIN	window.gbProductDetailInfo / __NEXT_DATA__	★★★★
Etsy	JSON-LD + variation JSON blocks	★★★
DHGate	window.__initial_state__	★★★
Wish	window.__data__ / __NEXT_DATA__	★★★★
Joom	window.__JOOM_DATA__	★★★
Banggood	window.productData	★★★
Gearbest	window.gbData	★★★
Any other site	JSON-LD + Open Graph meta (universal fallback)	—

---

Try it now

Paste this into the Apify Actor input form and click Run:

{
  "sources": [
    "https://www.aliexpress.com/item/1005006500000001.html",
    "https://www.etsy.com/listing/123456789/handmade-ceramic-mug",
    "https://www.temu.com/goods.html?goods_id=601099512345678"
  ],
  "targetCountry": "US",
  "targetCurrency": "USD",
  "maxConcurrency": 3
}

---

What makes this different

Variant-First extraction

Most scrapers only grab the displayed price. This actor reads the full SKU matrix embedded in each site's hydration JSON before JavaScript renders anything — so if a blue shirt has 5 sizes, you get 5 SKU entries, each with its own price, stock count, and variant-specific image.

"variants": [
  { "sku_id": "1,10", "color": "Blue", "size": "S", "sku_price": 12.99, "stock_status": "in_stock", "stock_quantity": 47, "variant_image": "https://..." },
  { "sku_id": "1,11", "color": "Blue", "size": "M", "sku_price": 12.99, "stock_status": "in_stock", "stock_quantity": 23, "variant_image": "https://..." },
  { "sku_id": "1,12", "color": "Blue", "size": "L", "sku_price": 13.99, "stock_status": "out_of_stock", "stock_quantity": 0, "variant_image": "https://..." }
]

Shadow Pricing captured

Temu and AliExpress show different prices based on selected variant. Because this actor reads the full price list from the hydration JSON — not the rendered DOM — every variant's exact price is captured, including discounts that only appear after selection.

Stealth browser with fingerprint injection

Anti-bot systems like DataDome and Cloudflare Bot Management are defeated through:

• Fingerprint injection — randomised navigator.webdriver, canvas hash, audio context, plugins list per request
• CDP timezone spoofing — browser timezone matches targetCountry
• Human-like behaviour — random mouse movement and delays before extraction
• Geo-targeted headers — Accept-Language and locale match the target country

AI-ready output

Every product includes a clean_text field: the description with all HTML tags stripped and entities decoded. Feed it directly into GPT-4 or Claude for competitor analysis, SEO copy generation, or product categorisation.

Geo-targeting & Currency

Set targetCountry: "DE" and the browser will appear to be a German user (language headers, timezone). Set targetCurrency: "EUR" and all prices are converted. Add a residential proxy from Germany for true local pricing.

---

Output format

One JSON object per product URL:

{
  "productId": "1005006500000001",
  "title": "2024 New Summer Women Dress Floral Print",
  "description": "<p>Material: 95% Cotton...</p>",
  "clean_text": "Material: 95% Cotton 5% Spandex. Available in 5 colors and 4 sizes.",
  "price": 14.99,
  "original_price": 29.99,
  "currency": "USD",
  "availability": true,
  "quantity": 312,
  "sku_count": 20,
  "variants": [
    {
      "sku_id": "Color:Red|Size:S",
      "attributes": { "Color": "Red", "Size": "S" },
      "color": "Red",
      "size": "S",
      "sku_price": 14.99,
      "original_price": 29.99,
      "stock_status": "in_stock",
      "stock_quantity": 48,
      "variant_image": "https://ae01.alicdn.com/kf/red-s.jpg"
    }
  ],
  "images": [
    "https://ae01.alicdn.com/kf/main.jpg",
    "https://ae01.alicdn.com/kf/detail1.jpg"
  ],
  "reviews_count": 1847,
  "rating": 4.8,
  "seller_id": "123456789",
  "seller_name": "Fashion Boutique Store",
  "shipping_cost": 0,
  "source_url": "https://www.aliexpress.com/item/1005006500000001.html",
  "site_label": "aliexpress",
  "country": "US",
  "scraped_at": "2026-03-04T12:00:00.000Z"
}

---

Input parameters

Parameter	Type	Required	Description
sources	string[]	✅	Product page URLs to scrape
targetCountry	string	—	ISO 3166-1 alpha-2 (e.g. US, DE, GB). Default: US
targetCurrency	string	—	Output currency (e.g. USD, EUR). Default: USD
proxyUrl	string	—	http://user:pass@host:port — use residential proxy for geo-accurate pricing
liveRatesUrl	string	—	Exchange rate API URL (e.g. open.er-api.com). Hardcoded rates used if omitted
maxConcurrency	integer	—	Parallel requests 1–20. Use 1–3 for Temu/AliExpress. Default: 3
priceSelector	string	—	CSS selector override for price (unsupported sites)
titleSelector	string	—	CSS selector override for title (unsupported sites)

---

Business intelligence use cases

Dropshipping & product sourcing

Find which Temu/AliExpress products have inventory on all sizes. Filter variants where stock_status === "out_of_stock" to avoid listing dead SKUs.

Competitor price tracking

Run daily on competitor Etsy listings. Compare sku_price per variant combination to spot when they run size-specific discounts.

LLM-powered product feed enrichment

Feed clean_text into GPT-4 to auto-generate localised product descriptions, SEO meta titles, or translate listings. No preprocessing needed.

Buy-Box analysis (Etsy / DHGate)

Multiple sellers list identical products. seller_id and shipping_cost let you identify the cheapest total-cost option and track which seller holds the Buy Box.

Pricing parity monitoring

Same product, different countries. Run with targetCountry: "US" and targetCountry: "DE" (plus matching residential proxies) to detect price discrimination across markets.

---

Adding a new site

The architecture is modular. To add support for a new store:

1. Create src/sites/mystore.js following this interface:

const MATCH_RE = /mystore\.com/i;

function matches(url) { return MATCH_RE.test(url); }

const LABEL = 'mystore';

function extractFromWindowVars(html) {
    // 1. Try to find window.__STORE_DATA__ or similar
    // 2. Return canonical product object, or null to fall back
    return null;
}

const CSS_SELECTORS = {
    title      : 'h1.product-title',
    price      : '.price-current',
    description: '#product-description',
    images     : '.product-gallery img',
};

module.exports = { matches, label: LABEL, extractFromWindowVars, CSS_SELECTORS };

2. Register it in src/sites/index.js — insert it before the generic entry.

That's it. The orchestration, variant normalisation, stealth browser, and currency conversion all work automatically.

---

Architecture overview

src/
├── main.js                  Apify Actor entry point — concurrency + currency
├── fetcher.js               Stealth Playwright browser + HTTP fallback
├── parser.js                Multi-strategy extraction pipeline
├── variant_extractor.js     Cartesian product SKU builder
├── stealth_browser.js       Fingerprint injection, anti-challenge helpers
├── geo_config.js            Country → locale/timezone/currency profiles
├── normalize.js             Price parsing + currency conversion
├── utils.js                 cleanText(), extractASIN(), safeJsonParse()
└── sites/
    ├── index.js             Site registry + URL resolver
    ├── aliexpress.js        window.runParams extractor
    ├── temu.js              __NEXT_DATA__ extractor
    ├── shein.js             gbProductDetailInfo extractor
    ├── etsy.js              JSON-LD + variation blocks extractor
    ├── dhgate.js            __initial_state__ extractor
    ├── wish.js              __data__ extractor
    ├── joom.js              __JOOM_DATA__ extractor
    ├── banggood.js          productData extractor
    ├── gearbest.js          gbData extractor
    └── generic.js           JSON-LD + Open Graph fallback

---

Legal & ethical use

This actor reads only publicly visible product information.

• Do not use authenticated sessions or bypass paywalls
• You are responsible for compliance with each website's Terms of Service
• Use residential proxies only through legitimate proxy providers
• Respect robots.txt and crawl rate limits (maxConcurrency: 1–3 for protected sites)

Q: How often should I run it?
A: Most teams run daily or weekly. Real-time monitoring isn't needed for DRAM pricing.

Q: Does it work with market indexes like DRAMeXchange?
A: Not yet. Market/index sites use different data formats. Planned for v1.1.

Q: Can I export the data?
A: Yes. All results are stored in JSON format and can be exported to CSV, Google Sheets, or via API.

---

Get started

1. Click Try for free in the Apify Console
2. Paste the demo input above
3. Click Run
4. See price data from real Mouser and Digi-Key products

Need help? Check the examples/ for more input samples.