Wildberries Product Search Scraper

Search public Wildberries product listings by keyword and return one ranked listing record per product. This Actor is for marketplace analysts, price-monitoring teams, sourcing researchers, and AI workflows that need a keyword in and clean Wildberries search cards out: NM-ID, public product URL, name, brand, supplier, current and original RUB price, discount, rating, review count, listed quantity, image count, source keyword, rank, destination, and collection time.

It queries Wildberries' public product-search JSON and saves only validated listing records. Product-detail descriptions, specifications, review threads, category URLs, and seller legal profiles belong in a follow-on Actor after this search pass.

Best fit for this Actor

• Start with a Russian or English product keyword when you need ranked Wildberries search listings rather than a full catalog crawl.
• Use maxResults as the spend cap: each saved product is one product-found event, plus Apify platform usage.
• Continue with a dedicated detail or review Actor when you need specifications, Q&A, or review text after this search pass.

Practical scenario

A sourcing analyst pastes ноутбук, keeps maxResults at 25, and optionally sets a rating floor. The dataset returns ranked public cards with sale and original RUB prices, discount, brand, supplier, rating, and the canonical product URL. The team can compare suppliers on the same keyword, then schedule the same input to watch rank and price movement. OUTPUT and RUN_SUMMARY say whether collection finished, stopped at the page cap, found no matches, or needed a later retry.

Quick start input

{
  "searchQueries": ["ноутбук"],
  "maxResults": 5,
  "sortBy": "popular"
}

searchQueries is required and accepts up to 10 distinct keywords. maxResults defaults to 50 and permits at most 1,000 records across the whole run. The Console quality sample uses 5 products so a first run stays bounded.

Input reference

Field	Type	What it controls
searchQueries	string[]	Product keywords or brand phrases, for example ноутбук. Required. Deduplicated. Not a category or product URL.
maxResults	integer	Overall saved-product cap, 1–1,000. Default 50.
maxPagesPerQuery	integer	Request-safety page cap per keyword, 1–50. Default 10. Reaching it before the natural end is PARTIAL.
destination	integer	Wildberries destination ID. Default -1257786 (Moscow).
sortBy	enum	Native search order: popular, rate, priceup, pricedown, newly, benefit.
minPriceRub / maxPriceRub	number	Optional post-search current-price filters in RUB.
minRating	number	Optional 0–5 post-search rating floor.
proxyConfiguration	object	Defaults to Apify Residential in Russia. Direct traffic is often rate-limited.

What data you receive

Each dataset item is one public listing product.

{
  "type": "wildberries_product",
  "productId": "12345678",
  "productName": "Ноутбук Example 14",
  "brand": "ExampleBrand",
  "supplier": "Example Supplier",
  "priceRub": 49209,
  "originalPriceRub": 67226,
  "discountPercent": 26.8,
  "currency": "RUB",
  "rating": 4.9,
  "reviewCount": 10,
  "imageCount": 14,
  "totalQuantity": 10,
  "productUrl": "https://www.wildberries.ru/catalog/12345678/detail.aspx",
  "sourceQuery": "ноутбук",
  "searchRank": 1,
  "destination": -1257786,
  "scrapedAt": "collection-time"
}

The default key-value store also contains:

• OUTPUT — outcome, message, saved count, warnings, and charged event counts.
• RUN_SUMMARY — input, request attempts, pages, filtering, duplicates, validation errors, and terminal diagnostics.

COMPLETE, PARTIAL, VALID_EMPTY, and INVALID_INPUT finish successfully. UPSTREAM_FAILED is the honest terminal state when valid work received no usable marketplace data after retries. Control rows stay out of the product dataset.

Use through the API

curl "https://api.apify.com/v2/acts/khadinakbar~wildberries-product-search-scraper/runs?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"searchQueries":["ноутбук"],"maxResults":25}'

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('khadinakbar/wildberries-product-search-scraper').call({
  searchQueries: ['ноутбук'],
  maxResults: 25,
  minRating: 4.5,
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items);

Use with AI agents through Apify MCP

Search public Wildberries listings for ноутбук, save at most 25 ranked products, and return name, product URL, current and original RUB price, discount, brand, supplier, rating, review count, and search rank. Inspect OUTPUT.outcome before treating an empty dataset as a real no-match.

Tell the agent to keep maxResults small on the first call, preserve productUrl as provenance, and read RUN_SUMMARY when the outcome is PARTIAL or needs a later retry. Current MCP client setup: https://mcp.apify.com

Connect the workflow

After a ranked search, continue with a Wildberries product-detail or review Actor when you need specifications or review text. For a non-Wildberries marketplace search in the same portfolio, start with Alibaba Listings Scraper when the job is B2B supplier cards instead of Wildberries consumer listings.

Pricing

This Actor uses Pay per event plus Apify platform usage. Open the live Pricing tab for current event names and amounts. maxResults is the predictable cap on product events before platform compute and proxy usage. A product is validated before it is persisted and charged, so only accepted listing rows receive the product event.

Best results

• Use Russian marketplace language when that matches shopper intent (ноутбук, беспроводные наушники).
• Keep the first run small (maxResults 5–25) and inspect OUTPUT before raising the cap.
• Leave the residential Russia proxy default unless you have a tested route. The public search JSON is the supported source; start there before HTML pages.

Builder's note

I found that Wildberries listing prices now live under sizes[].price.product / basic, and the search JSON sits at the top level as { products, total } rather than data.products. I built the parser to read both envelopes and to prefer reviewRating over the coarser integer rating bucket so a search row keeps a usable price and rating.

Responsible use

Use this Actor for lawful collection and analysis of public marketplace information. Follow applicable laws and Wildberries policies, and keep the workflow on public listing data you are authorized to collect. Wildberries is a trademark of its owner. This independent Actor is not affiliated with, associated with, or endorsed by Wildberries.