Americanas Product Scraper – Prices, Brands & Stock

Americanas Product Scraper

Watch the demo: https://youtu.be/U8GztLWKhZ0

Get product data from americanas.com.br, one of Brazil's largest online marketplaces, as clean rows: name, brand, price, list price, stock, seller, product URL, image and category. It reads the store's public catalog, so there is no login, no proxy and no browser. Only public product data is collected; no personal data.

Price: $2 per 1,000 products ($0.002 per product). Searches that return nothing are free.

Português: README.pt-BR.md

Use cases

• Price monitoring: track prices and discounts for your products or your competitors' on a schedule.
• Assortment research: see which brands and sellers are listed for a keyword or category.
• Stock checks: find out whether products are available.
• Market datasets: feed prices into Google Sheets, BI tools or an LLM.

Input

Field	Default	Description
searchTerms	–	Words to search, e.g. notebook, air fryer. Each term is scraped separately.
categoryUrls	–	Category pages, e.g. https://www.americanas.com.br/computadores-e-informatica/notebook (the /categoria/... form works too).
maxItems	100	Max products per search or category, 1–2500. Each product is one billable result.
sort	relevance	relevance, price_asc, price_desc, newest or best_selling.

At least one search term or category URL is required. Prices are in Brazilian reais (BRL); search terms work best in Portuguese (notebook, air fryer, tênis).

Example input:

{
  "searchTerms": ["air fryer"],
  "maxItems": 50,
  "sort": "price_asc"
}

Output sample

{
  "name": "Protetor Descartável Papel 50 Unidades Electrolux Air Fryer Antiaderente",
  "brand": "Electrolux",
  "sku": "3053207",
  "productId": "2866144",
  "price": 26.9,
  "listPrice": 27.9,
  "availability": true,
  "availableQuantity": 1,
  "seller": "Electrolux",
  "url": "https://www.americanas.com.br/protetor-descartavel-antiaderente-para-air-fryer-50-unidades-paela01-4720094973/p",
  "image": "https://americanas.vteximg.com.br/arquivos/ids/6587079/4720094981_1_xlarge.jpg?v=638751659805700000",
  "category": "Utilidades domésticas > Descartáveis > Assadeira Descartável",
  "source": "air fryer",
  "sourceType": "search",
  "scrapedAt": "2026-11-05T00:48:16Z"
}

Prices are in BRL. Export as JSON, CSV or Excel, call it by API, schedule it, or connect it to Make, Zapier or Google Sheets.

What you get

One dataset item per product:

Field	Description
name, brand	Product name and brand
productId, sku	VTEX product ID and the SKU (item) chosen for the price
price, listPrice	Current price and "from" price, in BRL
availability, availableQuantity	true if the chosen offer has stock; the quantity the store reports
seller	Seller of the offer (the store or a marketplace seller)
url, image	Product page and main image
category	Deepest category, e.g. Computadores e Informática > Notebook
source, sourceType	The search term or category path this item came from, and search or category
scrapedAt	UTC timestamp

Products are de-duplicated by productId inside each search or category. When a product has several SKUs or sellers, the first purchasable one is reported (in stock, price above zero); out-of-stock products are still returned, with availability: false.

Targets that return nothing, or that fail after retries, produce a status item (status, reason, source) instead. Status items are never billed.

Pricing (pay per event)

One billable event, product-scraped, for each product pushed to the dataset: planned price US$ 2.00 per 1,000 products (US$ 0.002 each). Status items are free. The Actor stops by itself when your maximum cost per run is reached.

Maintainer note: the event is configured in the Apify Console (Monetization → Pay per event), not in actor.json. Create the event product-scraped with price US$ 0.002. The code charges with Actor.push_data(item, "product-scraped"), which bills once per stored item and reports when the user's budget limit is hit.

How it works and how politely

• Calls GET /api/catalog_system/pub/products/search (50 products per page); categories are mapped to the store's category tree and queried with fq=C:/…/.
• One request at a time, about 1 second between pages, a real browser User-Agent.
• Retries with exponential backoff on HTTP 429 and 5xx and on network errors; other errors stop that target and are reported as a status item. The run continues with the next target.
• Product data only. No personal data is collected.

Limits

• The store caps a search at 2,500 results; narrow the term or use categories for more.
• Search relevance is the store's own and can include accessories (a search for notebook also returns notebook stands).
• Prices and stock are those shown to an anonymous visitor in Brazil; shipping costs are not included.

Local development

pip install -r requirements.txt pytest respx ruff
python -m pytest -q
ruff check .
python scripts/live_test.py   # real requests, writes docs/test-summary.md

Use of the data is your responsibility; check the site's terms and applicable law before you rely on it commercially.

More from Errand Code

This Actor is built and maintained by Errand Code. See the other tools: an AI visibility check (GEO audit), a Google Ads Transparency scraper, a Google Sheets add-on and MCP servers for AI agents.