Skip to content

1688 Search Scraper for Wholesale Listings

1688 is the wholesale side of Alibaba's ecosystem, and it sits behind bot detection that punishes naive request patterns.

OUTPUT FORMATSHTMLJSONMarkdown

This page explains how an API-based approach replaces hand-rolled proxy rotation and headless browser management with a single call, what data comes back, and how to size the work before you build. It is a product overview for teams running pricing intelligence, supplier vetting, or cross-border procurement feeds.

What the API does in one call

You send a GET request with your API key and the 1688 URL you want. The service renders JavaScript, rotates residential proxies, and returns rendered HTML ready to parse. Optional parameters extend that base: render_js for JavaScript execution, stealth_proxy for CAPTCHA-heavy targets, country_code for geo-routing, and ai_extract_rules for natural-language extraction. The same request can return parsed JSON when you supply extraction rules.

This matters because 1688's defenses are not a single wall. They combine fingerprinting, IP reputation, and client-side rendering. Handling each separately is a maintenance project; the API bundles them behind one endpoint so your team spends time on the data pipeline instead of the transport layer.

Fields you can expect from a 1688 product page

The page states that every 1688 product page returns title, tier pricing, MOQ, supplier name, seller rating, lead time, packaging dimensions, and the full image gallery. Treat that list as the target schema for a pricing or supplier-vetting table. A practical field map:

  • Pricing intelligence: title, tier pricing, MOQ, lead time.
  • Supplier vetting: supplier name, seller rating.
  • Logistics planning: packaging dimensions, lead time.
  • Merchandising: image gallery, title.

If a field is missing from a response, check whether the page variant you requested exposes it before assuming the API dropped it. 1688 layouts vary by category and listing type.

Three common pipelines and how they map

Most teams scraping 1688 fall into one of three patterns. Each maps to a different slice of the field list above.

Pricing intelligence on Chinese wholesale. Pull title, tier pricing, MOQ, and lead time for every SKU in your dataset. A weekly refresh cadence is enough to catch supplier price moves without hammering the target.

Supplier vetting. Collect seller ratings alongside supplier names to compare credibility before procurement or partnership decisions. Ratings are a signal, not a verdict; pair them with sample orders and compliance checks.

Cross-border procurement research. Ingest product feeds and packaging dimensions to model shipping efficiency and landed cost. Packaging data is often the difference between a viable and non-viable import.

Request controls and what each one fixes

The controls are the part most teams get wrong, because a failed request can look like a data problem when it is a routing problem. Use this map to diagnose:

  • render_js=true — the page needs client-side execution before content appears. The API runs it in a real headless browser and returns post-render HTML.
  • stealth_proxy=true — the target is blocking on fingerprint or IP reputation. This enables rotation, fingerprinting, and JS execution together.
  • country_code — content varies by geography. Routes through a residential IP in the named country.
  • extract_rules — you want parsed JSON. Map CSS selectors to field names.
  • ai_extract_rules — you want parsed JSON but prefer a natural-language spec over selectors.

A useful failure distinction: if the HTML comes back but your selectors find nothing, the page rendered a different layout — check the URL variant. If the request is blocked outright, the issue is transport, and stealth_proxy is the relevant control. If content differs from what you see in a browser, geo-routing is the likely gap.

Sizing a plan with the credit rules

Credits are consumed per request, and the cost depends on which features you enable. The supplied rules are: requests start at 1 credit each; JS rendering costs 5 credits per successful request; premium proxies cost 10 or 25 credits depending on whether JS rendering is also on. Credits are valid for one month and do not roll over.

A worked example using those rules: a weekly refresh of 10,000 product pages with JS rendering enabled costs 50,000 credits per run. Four runs a month is 200,000 credits, which sits between the Freelance and Startup tiers on the published plan table. Add premium proxies and the same workload moves to 1,000,000 credits per month at the 25-credit rate, which lands on Startup. Run the arithmetic against your own page count before choosing a tier; the plan table below is the reference for concurrency and included features.

Published plan reference

The table below reproduces the plan tiers, monthly prices, credit allowances, concurrency limits, and feature inclusions exactly as published. Prices are exclusive of VAT. Treat the figures as a dated reference and confirm current pricing before committing. The plan table is the authoritative source for concurrency and feature availability; the credit-cost rules above determine how far a given allowance stretches.

Setup path

The documented path is four steps. Create an account to get an API key and free starting credits. Install the SDK for your stack, or skip the SDK and use plain cURL. Send a GET request with your key and the 1688 URL. Add advanced parameters only when the base request does not return what you need.

SDKs are published for Python, Node.js, Go, PHP, Ruby, and Java, and the dashboard request builder generates snippets in Python, Node, Go, PHP, Ruby, or cURL. The request builder is the fastest way to confirm a working request before wiring it into a pipeline.

Rate limits and what to do about them

The API enforces concurrency limits per plan rather than per-target rate limits. If the target itself throttles aggressive patterns, the fix is on your side: space out requests or use multiple sessions. This is a different failure mode from a plan concurrency cap. A concurrency cap shows up as queued or rejected requests at your client; target throttling shows up as degraded responses from 1688. Diagnose which one you are hitting before changing plans.

Published plan reference
PlanMonthly priceAPI creditsConcurrent requestsPriority supportAccount managerTeam management
Hobby$19 /mo75,00025Not includedNot includedNot included
Freelance$49 /mo250,00050Not includedNot includedNot included
Startup$99 /mo1,000,000100Not includedNot includedNot included
Business$249 /mo3,000,000200IncludedIncludedIncluded
Business +$599 /mo8,000,000400IncludedIncludedIncluded
Enterprise 14$999 /mo14,000,000500IncludedIncludedIncluded
Enterprise 24$1,599 /mo24,000,000600IncludedIncludedIncluded
Enterprise 41$2,399 /mo41,000,000700IncludedIncludedIncluded
Enterprise 69$3,599 /mo69,000,000800IncludedIncludedIncluded
Enterprise 120$5,799 /mo120,000,000900IncludedIncludedIncluded

Prices are exclusive of VAT. All tiers include JavaScript rendering, rotating and premium proxies, geotargeting, and screenshots and extraction rules.

Frequently asked questions

Can I scrape 1688 search results without writing code?

Yes. The dashboard request builder lets you paste a URL, set parameters, and copy the generated snippet in Python, Node, Go, PHP, Ruby, or cURL. That snippet is a working request you can run as-is or drop into a pipeline.

How do I handle JavaScript when scraping 1688 search results?

Pass render_js=true on the request. The page runs in a real headless browser, the JavaScript executes server-side, and you get the post-render HTML back. If the content still does not appear, the page may need a different URL variant rather than more rendering.

Can I scrape 1688 search results from a specific country?

Yes. Pass country_code on the request to route through a residential IP in that country. This is useful when content varies by geography, which is common on marketplace pages.

What's the rate limit for scraping 1688 search results?

The API has concurrency limits per plan, not per-target rate limits. If the target throttles aggressive patterns, space out requests or use multiple sessions. Check whether you are hitting your plan's concurrency cap or the target's own throttling before changing tiers.

How do I get structured JSON from 1688 search results?

Two routes produce parsed JSON. The selector route takes extract_rules, where each CSS selector is paired with the field name you want it stored under. The prompt route takes ai_extract_rules, where you describe the fields you need in plain language instead of writing selectors. Selector rules stay predictable when the layout is stable; prompt rules are quicker to author when markup shifts. Either way, the response body arrives already parsed rather than as raw HTML.

Which programming languages can I use to scrape 1688 search results?

Any. SDKs are published for Python, Node.js, Go, PHP, Ruby, and Java, and plain cURL works too. Pick whatever your stack already uses; the request contract is the same across clients.