TL;DR
AliExpress renders its listings with JavaScript and rotates its CSS class names, so scrapers built on HTML selectors break within weeks. The reliable approach is pulling the JSON that AliExpress embeds in every search page, calling the hidden review API directly, and routing requests through rotating residential proxies so rate limiting never catches up with you. You get structured product data with plain Python and no headless browser for most targets.
AliExpress is one of the largest ecommerce platforms on the planet, and its product listings, prices, and reviews feed everything from dropshipping stores to market research pipelines. Getting that data out is harder than it looks because the site leans on JavaScript rendering, geo-based pricing, and aggressive rate limiting. In this article, we'll explore how to scrape AliExpress product data with Python, why most tutorials break within weeks, and how to keep your scraper running at scale.
Does AliExpress Have an Official API
Several popular scraping guides claim AliExpress has no API at all. That is wrong. The AliExpress Open Platform offers official endpoints for product details, affiliate links, and order management. The catch is access. You have to apply with a registered app, wait for manual approval, and the data is scoped to affiliate and dropshipping partners. Rate limits are tight, and many fields you actually want, like full review text or real-time stock, are not exposed. That gap is why most teams end up scraping the site directly.
How AliExpress Renders Product Data
AliExpress generates its CSS class names dynamically. A selector like price-default--current or list--gallery--C2f2tvm contains a hashed fragment that changes when the frontend redeploys, which is why scrapers built on those classes keep breaking. Some top-ranking tutorials still parse these exact classes, and their code quietly stopped matching anything.
The better target is the JSON AliExpress ships inside the page itself. Every search and category page embeds the full result set in a script tag variable called _init_data_. Pull that JSON out with one regex, and you get product IDs, titles, prices, ratings, and images as clean structured data. No selectors, no rendering, no breakage when a class name rotates.
Also Read: How to Do Web Scraping Without Getting Blocked
Scraping AliExpress Search Results with Python
Here is a working search scraper using plain httpx. No paid SDK required, unlike the code in most ranking tutorials. Route it through rotating residential proxies so each request exits from a different IP:
import httpx
import json
import re
proxy = "http://user:pass@gateway.proxyon.io:8000"
headers = {
"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36",
"Cookie": "aep_usuc_f=site=glo&c_tp=USD®ion=US&b_locale=en_US",
}
url = "https://www.aliexpress.com/w/wholesale-mechanical-keyboard.html"
resp = httpx.get(url, headers=headers, proxy=proxy, follow_redirects=True)
raw = re.search(r"_init_data_\s*=\s*{\s*data:\s*({.+}) }", resp.text).group(1)
fields = json.loads(raw)["data"]["root"]["fields"]
for item in fields["mods"]["itemList"]["content"]:
print(
item["productId"],
item["title"]["displayTitle"],
item["prices"]["salePrice"]["formattedPrice"],
)The aep_usuc_f cookie locks your session to the US storefront and USD pricing. Skip it, and AliExpress will guess a locale from your proxy IP, which makes your dataset inconsistent from one request to the next.
For pagination, append ?page=2 and so on. The embedded JSON includes totalResults and pageSize, so you can compute the page count from the first response instead of guessing.
Scraping Product Reviews Through the Hidden API
Reviews never touch the HTML. AliExpress loads them from a background endpoint at feedback.aliexpress.com, and you can call it directly:
product_id = "1005006717259012"
api = (
"https://feedback.aliexpress.com/pc/searchEvaluation.do"
f"?productId={product_id}&lang=en_US&country=US"
"&page=1&pageSize=20&filter=all"
)
resp = httpx.get(api, headers=headers, proxy=proxy)
data = resp.json()["data"]
print(data["totalPage"], "pages of reviews")
for review in data["evaViewList"]:
print(review["buyerCountry"], review["buyerTranslationFeedback"])This endpoint returns star distributions, translated review text, buyer country, and SKU info as raw JSON. It paginates cleanly through the page parameter, and it is far cheaper to hit than rendering product pages in a browser.
Matching Your Proxy Location to Your Target Market
AliExpress adjusts prices, shipping costs, and even product availability based on the visitor's location. If your requests exit from mixed countries, your price data is garbage. The fix is pinning both sides: set the region in the aep_usuc_f cookie and use proxies that exit from the same country.
Residential IPs work best here because AliExpress trusts them far more than datacenter ranges on product pages, where captcha challenges are common. For high-volume search scraping where speed matters more than trust, datacenter proxies handle the load at a lower cost per request. Product pages sometimes require JavaScript rendering when a captcha wall appears, in which case a headless browser like Playwright behind a residential IP gets you through.
Also Read: How to Build a Price Monitoring Scraper With Proxies
FAQ Section
Is it legal to scrape AliExpress?
Scraping publicly available product data is generally legal, but AliExpress prohibits automated access in its terms of service, and laws vary by jurisdiction. Keep request rates respectful, avoid personal data, and get legal advice if you plan to redistribute what you collect.
Why does my scraper return different prices than my browser?
AliExpress localizes prices by IP, currency, and region settings, so your browser session and your scraper are almost certainly resolving to different storefronts. Pin the aep_usuc_f cookie and match your proxy country to fix it.
Do I need a headless browser to scrape AliExpress?
Not for search and category pages, since the data sits in embedded JSON. Product pages are protected more aggressively and sometimes serve CAPTCHA challenges, so keep a browser fallback ready for those, ideally behind residential IPs.
What data can you scrape from AliExpress?
Search and category pages give you product IDs, titles, sale and original prices, ratings, order counts, and image URLs through the embedded JSON. Product pages add specifications, shipping options, and seller details, while the review API covers star distributions and full review text. That combination is enough for price monitoring, catalog building, and supplier research.
Can I scrape AliExpress without proxies?
You can for a handful of requests, but AliExpress rate limits by IP and starts serving captcha challenges fast. A single IP pulling hundreds of pages is an obvious bot signature. Rotating IPs spreads the traffic so each address stays under the radar.
How do I change the currency or country in my scraped data?
Everything runs through the aep_usuc_f cookie. Set c_tp to your currency code, region to the country, and b_locale to the language, then match your proxy exit country to the same region. Mismatched settings are the most common reason scraped prices look wrong.
Final Thoughts
Scraping AliExpress comes down to picking targets that do not move. The embedded JSON and the review API stay stable while CSS classes rotate underneath every tutorial that depends on them. Pair those endpoints with rotating residential IPs matched to your target market, and the blocks mostly disappear.
