Costco Scraper

Extract Costco products, prices, reviews, categories, and warehouse locations across eight markets as rows you can export.

• Search products or browse categories and Hot Buys.
• Get product identifiers, prices, ratings, availability, and images when shown.
• Collect reviews, category entries, or warehouse addresses as separate row types.
• Choose the United States, Canada, Mexico, United Kingdom, Australia, Japan, South Korea, or Taiwan.
• Pay only for rows delivered: $3.00 per 1,000 rows on the Free plan, plus a start fee.

What does Costco Scraper do?

Costco Scraper collects publicly available shopping data from costco.com, costco.ca, costco.com.mx, costco.co.uk, costco.com.au, costco.co.jp, costco.co.kr, and costco.com.tw. Use one target per run for price tracking, assortment research, review analysis, or location lookup.

• Product listings: search with keyword, browse with categoryUrl or categorySlug, or choose a topic.
• Product details: use productUrl or itemId.
• Reviews: choose productReviews; each review is a separately billed row, not an automatic addition to a product row.
• Locations and categories: choose warehouseLocations or categoryTree.
• Optional enrichment: enrichDetails is off by default. It adds available product details to existing listing rows, rather than creating extra rows.

This is an unofficial tool, not affiliated with Costco. It does not place orders or access private member accounts.

What data can you extract from Costco?

Fields depend on the selected mode and what Costco provides. Missing values may be absent or null. rowType distinguishes products, reviews, warehouses, and categories.

Product rows

Field	Description	Example
rowType	Entity type	product
itemId	Costco item number	4000376434
sku	Additional product identifier, when supplied	null
title	Product name	Paper towels, 12 rolls
url	Product link	https://www.costco.com/p/-/ninja-crispi-portable-air-fryer/4000376434
price	Displayed price, not a checkout quote	119.99
currency	Currency associated with the price	USD
brand	Brand shown for the product	Ninja
rating	Average product rating	4.7
reviewCount	Number of ratings or reviews shown	125
availability	Availability information when supplied	InStock
images	Product image links, when available	[]
pdp	Additional details on enriched listing rows	{"brand":"Ninja"}
browseCountry	Selected market	US
sourceSurface	Type of page supplying the row	keywordSearch
scrapedAt	Collection timestamp	2026-10-24T00:00:00Z

Review rows

Field	Description	Example
rowType	Entity type	review
reviewId	Review identifier	example-review-1
itemId	Reviewed item's number, when supplied	10024438
title	Review heading	Useful size
text	Review content when provided under this field	Fits the shelf well.
rating	Individual review score	4
author	Public reviewer display name	Jane Doe
submissionTime	Review publication timestamp	2026-10-20T12:00:00Z

Warehouse rows

Field	Description	Example
rowType	Entity type	warehouse
name	Warehouse name	Example warehouse
url	Warehouse or locator link	https://www.costco.com/warehouse-locations
address	Available street, locality, region, postal code, and country	{"locality":"Seattle","country":"US"}
phones	Public warehouse phone entries	[{"value":"+1 555 0100","type":"phone"}]

Category rows

Field	Description	Example
rowType	Entity type	category
name	Category name	Grocery & Household
slug	Category path identifier	grocery-household
url	Category link	https://www.costco.com/grocery-household.html
depth	Category level, when available	1
parentId	Parent category identifier, when supplied	null
path	Category hierarchy, when supplied	["grocery-household"]

How to scrape Costco

1. Open Costco Scraper in Apify Console.
2. Choose browseCountry and fill one target, such as keyword.
3. Set maxResults to 3 for a small first run.
4. Click Start and wait for completion.
5. Open the dataset and download your rows or read them through the API.

Empty input starts the form's example: paper towels in CA (Canada), limited to 3 rows. This is a billed example, not a free preview. With mode set to auto, filling one target selects its matching mode. Explicit warehouseLocations and categoryTree runs need no target and do not become examples.

Keyword search

Fill keyword, such as paper towels. Set browseCountry for the desired market; use words appropriate to that market.

Category browse

Fill categoryUrl or categorySlug. Full links must match browseCountry. A US example is grocery-household.

Specialty browse

Set topic to hotBuys, whatsNew, warehouseOnly, or grocery. Topic availability varies by market.

Product detail

Fill productUrl or itemId. This selects productDetail automatically and returns details for one product, not a list of unrelated products.

Product reviews

Select mode: productReviews explicitly, then fill productUrl or itemId. Without this mode, a product target selects product details instead.

Warehouse locations

Select mode: warehouseLocations. Add zip to request a postal-code search where supported. Results are warehouse entries, not product-level stock guarantees.

Category tree

Select mode: categoryTree to get category entries. The returned list is not a guarantee of an exhaustive or fully current catalog.

What to fill in

Field	What to enter	Default
mode	auto or one of the seven modes above	auto
browseCountry	US, CA, MX, UK, AU, JP, KR, TW	US
keyword	One search phrase	None; form sample paper towels
query	Older name for keyword	Empty
productUrl	Product link matching the market	Empty
itemId	Product item number	Empty
slug	Optional product-name path for an item number	Empty
categoryUrl	Category link matching the market	Empty
categorySlug	Category path without the domain	Empty
topic	One of the four specialty topics	Empty
maxResults	Row limit, from 1 to 500	70; form sample 3
zip	Postal code for location context	Empty
warehouse	Warehouse identifier for location context	Empty
dept	Department filter for supported keyword searches	Empty
sort	relevance, priceAsc, priceDesc, newest, or rating	Site order
currency	Optional currency setting; not currency conversion	By market
page	Starting page, from 1 to 200	1
maxPages	Page cap per part of up to 70 rows, from 1 to 50	No explicit cap
enrichDetails	Add available details to listing rows	false
enrichLimit	Listings to enrich per part, from 0 to 50	5

Keyword search:

{
  "keyword": "paper towels",
  "browseCountry": "US",
  "maxResults": 3
}

Category browse:

{
  "categorySlug": "grocery-household",
  "maxResults": 3
}

Specialty browse:

{
  "topic": "hotBuys",
  "maxResults": 3
}

Product detail:

{
  "productUrl": "https://www.costco.com/p/-/ninja-crispi-portable-air-fryer/4000376434",
  "maxResults": 1
}

Product reviews:

{
  "mode": "productReviews",
  "itemId": "10024438",
  "maxResults": 3
}

Warehouse locations:

{
  "mode": "warehouseLocations",
  "zip": "98101",
  "maxResults": 3
}

Category tree:

{
  "mode": "categoryTree",
  "maxResults": 3
}

maxItems and limit are accepted aliases for maxResults. Unknown fields alone do not fail input validation: close misspellings are corrected where possible; other unknown fields are ignored with a note.

When targets conflict, precedence is productUrl, categoryUrl, itemId, categorySlug, topic, then keyword or query. Targets belonging to another mode are ignored, not collected as additional inputs. Use separate runs for separate targets.

The INPUT_REPORT record in the default key-value store records input interpretation, corrections, ignored fields, and example use. Invalid targets or market links stop before row collection with INVALID_INPUT: and a field-specific explanation. No rows are charged for input-validation failures; the start fee still applies.

Output example

Illustrative product row; prices and availability are not current offers:

{
  "rowType": "product",
  "itemId": "4000376434",
  "sku": null,
  "title": "Ninja Crispi 5-in-1 Portable Glass Air Fry Cooking System",
  "url": "https://www.costco.com/p/-/ninja-crispi-portable-air-fryer/4000376434",
  "price": 119.99,
  "currency": "USD",
  "rating": 4.7,
  "browseCountry": "US",
  "sourceSurface": "keywordSearch"
}

Export the dataset as JSON, CSV, Excel, XML, RSS, or an HTML table. Keep identifiers as text when importing into a spreadsheet.

Why use this Costco scraper?

• Pay for delivered rows. Product, review, category, and warehouse rows use the same per-row price.
• Keep delivered data. Rows already written to the dataset remain available after a stop. Recoverable partial results can be delivered; a stop before delivery may leave no rows.
• Understand failures. Input errors explain what to fix; RUN_SUMMARY and, on failure, ERROR_SUMMARY provide run information.
• Keep missing data honest. Unavailable fields stay empty rather than becoming invented prices or contact details.
• Compare markets deliberately. Select a market per run and keep browseCountry with your exports.
• No proxy setup. You do not need to set up proxies.

How much does it cost to scrape Costco?

The Actor uses pay-per-event pricing. At the default 512 MB memory, the start fee is $0.005: one start event per GB, with a minimum of one event. Platform usage is included in these prices.

Apify plan	Start fee at 512 MB	Price per row	Price per 1,000 rows
Free	$0.005	$0.003	$3.00
Starter	$0.005	$0.003	$3.00
Scale	$0.005	$0.0026	$2.60
Business	$0.005	$0.0022	$2.20
Platinum	$0.005	$0.002	$2.00
Diamond	$0.005	$0.0016	$1.60

• maxResults applies to the single input in a run, defaults to 70, and accepts up to 500. Searches, category listings, specialty listings, and reviews can continue in parts of up to 70 rows.
• Category trees and warehouse lists return at most 70 rows. Availability and page limits may reduce the result count.
• Free Apify plan users are limited to 200 rows per run and 10 runs of this Actor.
• Maximum cost per run (maxTotalChargeUsd in the API) limits the whole run. A budget too small for one row can leave an empty dataset after the start fee.

Example: one search delivering 5 rows on the Free plan at 512 MB costs 5 × $0.003 + $0.005 = $0.020. Requesting 5 rows does not guarantee 5 matches; row charges follow delivery.

What happens when the run timeout is too short

A short timeout can cause the run to pause instead of failing while results are pending. It shows ABORTED with a message like "Queued on the collection backend; this run will resume automatically when results are ready (about Ns)."

When ready, the same run resumes automatically, delivers into the same dataset, and finishes as SUCCEEDED if collection succeeds. A resume can report "Resumed: delivering N rows." Already delivered rows are skipped, so they are not delivered or charged twice.

This can happen at most 3 times per run. Keep the default timeout of 3600 seconds to reduce interruptions. Automatic resumption is not a promise that every manually aborted or failed run will recover.

The synchronous API endpoints run-sync and run-sync-get-dataset-items wait at most 300 seconds and can return an empty dataset while a run is paused. For longer work, start asynchronously and read the dataset after SUCCEEDED, or use an ACTOR.RUN.SUCCEEDED webhook.

Use with AI agents and the API

Apify MCP server

Add this remote tool in a compatible client such as Claude, ChatGPT, Cursor, or VS Code:

https://mcp.apify.com?tools=lentic_clockss/costco-scraper

Sign in through Apify OAuth, or send Authorization: Bearer <YOUR_APIFY_TOKEN>. The tool returns run and dataset IDs; the agent can read rows with get-dataset-items. Clients without remote support can use npx @apify/actors-mcp-server --tools lentic_clockss/costco-scraper with APIFY_TOKEN set locally.

Example prompt: "Search Costco US for paper towels, return at most 3 products, and show titles, prices, and links."

Apify API

from apify_client import ApifyClient

client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("lentic_clockss/costco-scraper").call(
    run_input={"keyword": "paper towels", "maxResults": 3}
)
for row in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(row)

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: '<YOUR_APIFY_TOKEN>' });
const run = await client.actor('lentic_clockss/costco-scraper').call({
  keyword: 'paper towels', maxResults: 3
});
const { items: rows } = await client.dataset(run.defaultDatasetId).listItems();
console.log(rows);

For asynchronous HTTP access, send the input as JSON to POST https://api.apify.com/v2/acts/lentic_clockss~costco-scraper/runs with your bearer token. After successful completion, read GET https://api.apify.com/v2/datasets/<datasetId>/items?format=json. If a waiting client returns an ABORTED run, apply the timeout guidance above rather than treating its dataset as final.

Integrations

Use Apify Schedules for recurring runs. Connect results through the Integrations tab to Google Sheets, Make, Zapier, n8n, Slack, Airtable, or your own webhook.

FAQ

Is it legal to scrape Costco?

Public availability does not remove your responsibilities. Check applicable terms and data-protection law for your use. Processing personal data, including reviewer names, requires a lawful basis. This Actor does not grant permission on Costco's behalf.

Why did my run end as ABORTED?

Read the status message. A queued run may resume automatically as described above. A manual abort or other interruption is different; check the run details before restarting and creating another billable run.

How many results can I get?

Set maxResults up to 500 for one input. The Free plan caps each run at 200 rows; warehouse and category-tree modes stop at 70. Source availability, page limits, and your spending limit can reduce delivery.

Why are some fields empty?

Not every page exposes a price, rating, image, phone number, or availability value. Listing enrichment may fill some fields, but does not guarantee them. An empty field is not a zero price or confirmed lack of stock.

Why do itemId and sku differ?

itemId identifies the item in the product link. sku is another identifier supplied with product details and may differ. Keep both when provided; neither is guaranteed on every row.

Can I send several keywords or links together?

No. Target fields accept single values. Competing target fields select one mode rather than creating a batch. Start separate runs for separate keywords, products, categories, or markets.

Are Same Day, Pharmacy, and China covered?

No. Same Day, Business Delivery, Pharmacy, Travel, and Photo are outside scope, as is China. Supported markets refer to Costco shopping sites, not all Costco services.

Why did a successful run return no rows?

A search with no matches can finish successfully with an empty dataset. Review the input and RUN_SUMMARY. Unavailable review data or collection failures are not a guarantee that the product has no reviews.

Limitations

• Market selection does not guarantee every mode, topic, department filter, or sort order works identically across countries.
• Prices and availability reflect exposed information, not member-specific discounts, checkout totals, or guaranteed warehouse inventory.
• Reviews and product-detail enrichment may be incomplete. Enriched listing details can appear inside pdp.
• Category entries may be incomplete or outdated; the category-tree mode is not a full-catalog product download.
• One target is processed per run. There is no documented date-window input or guaranteed automatic search-word translation.

Related scrapers

More e-commerce and marketplace scrapers from the same publisher:

• Walmart Product Scraper
• Temu Product Scraper
• Allegro Scraper
• TikTok Shop Product Scraper
• Amazon Search, Category & Best Sellers Scraper
• Coupang Product Scraper
• MercadoLibre Scraper
• Rakuten Japan Scraper

Browse all of our scrapers at apify.com/lentic_clockss.

Support

Report a problem or request a field through the Issues tab of this Actor's page. Include the run ID and the input you used.