Amazon India Product Scraper

Scrape public Amazon.in product data by search query, product URL, ASIN, search URL, or category URL. The actor uses bounded HTTP requests and source-specific HTML/JSON-LD parsing. It does not use a browser, stealth fingerprinting, CAPTCHA solving, sign-in, or checkout access. Optional Apify/custom proxy input is honored, with at most two retries and a 10 MB response limit.

Extracted fields

Product records can include:

• Title, ASIN, canonical/product URL, brand, breadcrumbs, and source/input provenance
• Current/list price with currency, stock and availability text, delivery-related public text when present
• Stars, review count, rating breakdown when exposed, public review link, and optional review cards
• Features, description, technical attributes/product overview, gallery/high-resolution images, and video/media metadata
• Direct HTTP status/content type, access/fetch status, stable record ID, field coverage, and data-quality metadata

The dataset contains product rows only. A requested ASIN is stored only when the response title and ASIN match that exact target. Blocks, mismatches, invalid input, and extraction failures are written to the bounded OUTPUT summary; a run with zero verified products fails closed with an empty dataset.

Input

Field	Type	Default	Description
mode	string	search	search, productUrl, searchUrl, categoryUrl, or asin
searchQuery	string	laptop	Search phrase for search mode
asins	array	—	Ten-character ASINs for asin mode
productUrls	array	—	Public Amazon.in product URLs for productUrl mode
searchUrl	string	—	Public Amazon.in search URL
categoryUrl	string	—	Public Amazon.in category/bestseller URL
maxItems	integer	10	Maximum product records, from 1 through 100
deepProductScraping	boolean	true	Legacy flag retained; direct mode always verifies product pages
productReviews	boolean	false	Parse up to 10 public review cards when present
requestDelayMs	integer	600	Delay between serialized requests
requestTimeoutMs	integer	60000	Direct HTTP timeout per request
maxRetries	integer	1	Bounded retries per request, from 0 through 2
proxyConfiguration	object	—	Optional Apify/custom proxy configuration used for every request

The legacy headless field remains accepted but is ignored because no browser is launched. When proxyConfiguration is omitted, requests are direct; when supplied, the configured proxy is used and reported in the run/output metadata.

Billing follows the Actor's current pay-per-event configuration: one apify-default-dataset-item event per verified product row. Failed or blocked runs keep the dataset empty and therefore create no result-item charge.

Example search input

{
  "mode": "search",
  "searchQuery": "laptop",
  "maxItems": 3,
  "productReviews": false,
  "requestDelayMs": 0
}

Example ASIN input

{
  "mode": "asin",
  "asins": ["B0EXAMPLE1"],
  "maxItems": 1,
  "productReviews": true
}

Output and limitations

Normal records expose stable fields such as title, url, asin, price, listPrice, inStock, inStockText, brand, stars, reviewsCount, features, attributes, thumbnailImage, galleryThumbnails, sourceUrl, httpStatus, fieldCoverage, and scrapedAt, plus optional public reviews and additional product metadata.

Amazon markup, availability, prices, and review sections change frequently. Missing source sections are omitted from output rather than represented by null or empty values. A blocked run is never presented as a successful product result.

Local testing

Create or update storage/key_value_stores/default/INPUT.json, then run:

apify run

Useful checks:

apify validate-schema
npm run lint
node validate-datasets.js

Inspect storage/datasets/default/ and storage/key_value_stores/default/OUTPUT.json. Local storage is not uploaded automatically.