Costco Scraper

Extract product data from Costco.com — prices, member prices, descriptions, ratings, images, and availability. Paste in product detail page URLs.

What it does

Point it at Costco.com product pages and it returns structured product records: item number, title, regular price, member price, description, rating, review count, availability, images, and category breadcrumb.

Costco's Akamai bot protection is handled for you.

Inputs

Field	Type	Description
startUrls	array	Costco product detail page URLs (https://www.costco.com/p/-/<slug>/<itemNumber>). Category/browse URLs are accepted as a best effort.
maxItems	integer	Cap on total records (default: 100)

If startUrls is empty the actor exits with an error. Keyword search is not supported — Costco's search endpoint is not reachable through the unblocker.

Output

Each record in the dataset:

{
  "item_number": "4000140441",
  "product_url": "https://www.costco.com/p/-/kutano-commercial-1-door-merchandiser-refrigerator/4000140441",
  "title": "Kutano Commercial 1-Door Merchandiser Refrigerator",
  "price": "$1849.99",
  "member_price": "",
  "description": "Kutano Commercial 1-Door Merchandiser Refrigerator, Stylish Black Laminated Vinyl...",
  "rating": "4.09",
  "review_count": "",
  "availability": "",
  "images": "https://bfasset.costco-static.com/...",
  "specifications": "",
  "categories": "",
  "scraped_at": "2026-06-24T19:37:21.596Z"
}

How it works

• Product detail URLs (/p/-/<slug>/<itemNumber>) go straight to the product handler, which parses Costco's Next.js RSC payload for the product data.
• Category/collection URLs are treated as listing pages: product links are collected from the rendered HTML and each product page is then fetched.

Bot protection is handled transparently, nothing to subscribe to.

Limitations

• Member pricing is only populated when Costco exposes it without authentication.
• Specifications table extraction depends on the page rendering consistently; some product types (electronics, tires) use non-standard markup.
• Keyword search is not supported. Provide product detail URLs directly.
• Category pages are large and slow through the unblocker, and link density varies by category — direct product URLs are the most reliable input.