AliExpress Scraper

Scrape AliExpress search results, categories, products, reviews, sellers, and discovery feeds without maintaining anti-bot headers, signatures, cookies, or browser sessions.

The Actor accepts AliExpress URLs with a two-letter country subdomain. Each request is processed independently and produces one row in the default dataset.

Quick start

{
  "requests": [
    {
      "url": "https://sg.aliexpress.com/w/wholesale-wireless-earbuds.html?page=1"
    }
  ]
}

A run accepts 1 to 100 requests. The Actor waits for every request to reach a terminal state before publishing results and charging finalized events.

Country routing

Every submitted URL must use a two-letter ISO country prefix:

https://${countryCode}.aliexpress.com/...

For example, use sg.aliexpress.com for Singapore or gb.aliexpress.com for the United Kingdom. The prefix selects the task country and proxy location. Do not submit www.aliexpress.com, because it does not identify a task country. The selected country must have a configured proxy.

Input

Shared options apply to every request. A value inside an individual request overrides the shared value for that request.

• requests - Required array containing 1 to 100 AliExpress requests.
• emulateMobileDevice - Emulate mobile web traffic. Default: false.
• language - Optional request language.

Per-request fields:

• url - Required AliExpress URL with a two-letter country subdomain.
• method - GET. Default: GET.
• Any shared option above - Overrides that option for this request.

cleanResponseBody is not supported. Unknown request options are passed to the backend for forward compatibility.

Product search and Seller products options

These options affect only Product search and Seller products requests, whether supplied at the input or request level:

• productListing_crawlNextPages - Crawl subsequent listing pages.
• productListing_crawlNextPages_maxPages - Maximum listing pages.
• productListing_crawlNextPages_maxUniqueProducts - Maximum unique products.
• productListing_crawlNextPages_stopWhenNewPageOnlyContainsDuplicates - Stop on a page containing only products already collected. Default: true.
• productListing_crawlNextPages_stopWhenNoMorePagesIsTrue - Stop when AliExpress reports no more pages. Default: true.

Supported AliExpress APIs

Product search by keyword

https://${countryCode}.aliexpress.com/w/wholesale-${keywordSlug}.html?page=${page}

{
  "requests": [
    { "url": "https://sg.aliexpress.com/w/wholesale-wireless-earbuds.html?page=1" }
  ]
}

Keyword products example

Product search by category

https://${countryCode}.aliexpress.com/category/${categoryId}/${categorySlug}.html?page=${page}

{
  "requests": [
    { "url": "https://sg.aliexpress.com/category/200003482/women-clothing.html?page=1" }
  ]
}

Category products example

Seller products

https://${countryCode}.aliexpress.com/store/${storeId}?page=${page}

The page parameter selects seller products. Without page, the URL selects seller details.

Seller products example

Product details

https://${countryCode}.aliexpress.com/item/${productId}.html

{
  "requests": [
    { "url": "https://sg.aliexpress.com/item/1005011672102938.html" }
  ]
}

Product detail example

Product reviews

https://${countryCode}.aliexpress.com/pdp/review/getReviewList?itemId=${productId}&page=${page}&filter=${filter}&sort=complex_default

Supported review parameters:

• page or pageNo - One-based page. Default: 1.
• filter - all or image. Default: all.
• sort - Only complex_default is supported.
• pageSize - Use 10 for all or 6 for image when supplied.

Star filters and alternative sort modes are not supported.

Product reviews example

Seller details

https://${countryCode}.aliexpress.com/store/${storeId}

Do not add page to a seller-detail URL.

Seller detail example

Category navigation

https://${countryCode}.aliexpress.com/h5/mtop.lazada.guided.shopping.categories.categorieslpcommon/1.0/

This endpoint extracts homepage category navigation for the selected country, not a complete category tree.

Category navigation example

Keyword listing

https://${countryCode}.aliexpress.com/fn/seo-alphabet/index?channel=popular&pageNo=${page}

The channel value must be popular.

Keyword listing example

Unsupported APIs

AliExpress does not support:

• Seller promoted products.
• Product Reviews V2. Use Product Reviews instead.
• Seller listing through sitemap-sellers.xml.

Unsupported requests are rejected before execution and are not billed.

Output

The default dataset contains one row per input request:

{
  "marketplaceRunUuid": "...",
  "taskUuid": "...",
  "url": "https://sg.aliexpress.com/item/1005011672102938.html",
  "status": "SUCCESS",
  "responseBody": {}
}

Output fields:

• marketplaceRunUuid - Marketplace run ID.
• taskUuid - Task ID.
• url - Original requested URL.
• status - Terminal request status.
• responseBody - Native AliExpress response for a successful request.
• error - Error description for an unsuccessful request.

Rows remain in the same logical request order. One failed request does not hide successful rows from the same batch.

Related Actors

• Lazada Scraper - Scrape Lazada searches, products, ratings, sellers, and categories.
• Shopee Scraper - Scrape Shopee Web searches, products, ratings, shops, categories, and recommendations.

Pricing and charging

Pay as you go with no subscription or monthly minimum:

• Successful AliExpress request: $0.00272 per request, or $2.72 per 1,000 successful requests.
• Proxy usage: $0.00013 per started block of 100 proxy units.
• Actor start: $0.00005 per GB of Actor memory, with a minimum of one event.
• Dataset result: $0.00001 per result written to the default dataset.

This is a PAY_PER_EVENT Actor. It charges one csea-scrape-aliexpress event per successful AliExpress request and finalized csea-proxy-unit events for proxy usage. Failed or cancelled requests are not charged as successful AliExpress requests. Use Apify's maximum total charge setting to bound an execution's cost.

Graceful aborts and retries

A graceful Apify abort cancels known scraping tasks and waits for backend finalization. Interrupted Actor executions persist their marketplace-run, charging, and output state so they can resume without intentionally duplicating charges or dataset rows.

Operational guidance

• Treat response fields and live totals as marketplace-controlled data.
• Monitor success rates during major sales and high-traffic periods.
• Prefer bounded listing crawls over unbounded multi-page requests.