Amazon Product Scraper — search results from 11 marketplaces as clean JSON

Extract Amazon search results — ASIN, title, current price, crossed-out list price and the discount between them, star rating, review count, delivery promise, badges and a tracking-free product link — as flat JSON rows for price monitoring, market research and catalogue work. Search any keyword on any of eleven Amazon marketplaces, with price bands, Prime filtering and sorting.

Live health record: lowlanddata.com/status/amazon-product-scraper - 30-day success rate and daily canary results, failures included. Full input/output reference: tool page.

No personal data, ever. Products are commercial objects and come back whole. Marketplace sellers do not: a third-party seller is frequently a sole trader, which makes their name personal data, so there is no seller field and there never will be. Reviews are counted, never quoted — no reviewer names, no review text.

Quick start (30 seconds)

1. Type what you would type into Amazon's search box into keywords.
2. Pick a marketplace — prices and availability differ per country.
3. Click Start, then open the dataset's Overview tab or Export as CSV/Excel/JSON.

Optional: minPrice/maxPrice, primeOnly, sortBy, and skipSponsored to drop paid placements. Everything works on a daily Schedule.

What you can build with it

• Price monitoring. Schedule daily on your own ASINs or your category, and watch price and discountPercent move. Rows are flat, so a spreadsheet or a database takes them as-is.
• Competitor and category research. Who ranks for your keyword, at what price, with how many reviews — the whole first page as data instead of screenshots.
• Repricing inputs. price, listPrice and discountPercent per ASIN across marketplaces, which is what a repricing rule actually needs.
• Assortment gap analysis. Compare the same keyword across amazon.de, .co.uk and .fr to see what sells where.

What you get

One JSON object per product:

{
  "asin": "B07QPS7L6Y",
  "title": "RAVIAD USB C Kabel [4Pack 0.5M 1M 2M 3M] Ladekabel Schnellladekabel",
  "url": "https://www.amazon.de/dp/B07QPS7L6Y",
  "marketplace": "de",
  "price": 6.79,
  "currency": "EUR",
  "listPrice": 9.99,
  "discountPercent": 32,
  "rating": 4.6,
  "reviewCount": 23993,
  "imageUrl": "https://m.media-amazon.com/images/I/81ZJli188JL._AC_UY218_.jpg",
  "deliveryText": "GRATIS Lieferung Di, 16. Sep",
  "isSponsored": false,
  "isAmazonsChoice": true,
  "isBestSeller": false
}

With scrapeDetails: true, rows additionally carry availability, featureBullets, category, bestSellerRank and bestSellerCategory — and the price is refreshed from the product page, which is fetched later than the search tile.

Links are canonical — amazon.xx/dp/ASIN — with Amazon's click-tracking blob removed, so they stay stable and shareable.

How much does it cost?

Pay per product row plus a small start fee — no subscription. Free-plan users can try it capped at 5 products.

scrapeDetails is charged separately, and only when a detail actually arrives. That is not a pricing trick, it is the cost: one search page is 276 KB and carries sixteen products, while one product page is 443 KB and carries one — about 25x the traffic per item. Callers who do not need stock and Best Sellers Rank never pay for them, and a detail fetch that fails is not charged at all; the search row still ships.

Platform usage is billed by Apify; Amazon needs a residential route, so runs cost a little more than a plain HTML site.

Not technical? Let your AI assistant set it up

I want to use the Amazon Product Scraper (lowlanddata/amazon-product-scraper) on Apify.

What I want: [E.G. "track the price of wireless keyboards under 50 EUR on amazon.de, daily"]

Guide me to:
1. Propose input values: keywords, marketplace, minPrice, maxPrice,
   primeOnly, sortBy, skipSponsored and maxItems.
2. Explain how to run it in the Apify Console and export CSV.
3. Suggest a daily Schedule so I can watch prices move over time.

Input reference

Field	Type	Default	Notes
keywords	string	usb cable	What you would type into Amazon's search box.
marketplace	string	de	de, uk, fr, it, es, nl, se, pl, be, us, ca.
minPrice	integer	—	Lower bound in the marketplace's currency.
maxPrice	integer	—	Upper bound in the marketplace's currency.
primeOnly	boolean	false	Only Prime-eligible products.
sortBy	string	relevance	relevance, price-asc, price-desc, newest, reviews.
skipSponsored	boolean	false	Drop paid placements and keep organic results only.
maxItems	integer	200	Stop after this many products.
proxyConfiguration	object	residential	Amazon blocks datacenter addresses quickly — keep this as it is.

Scheduling and API use

curl -X POST "https://api.apify.com/v2/acts/lowlanddata~amazon-product-scraper/runs?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"keywords": "wireless keyboard", "marketplace": "de", "maxPrice": 50, "maxItems": 100}'

Use it with AI agents (MCP)

Run it directly from Claude, Cursor and other MCP clients: apify.com/lowlanddata/amazon-product-scraper/api/mcp.

Is it legal to scrape Amazon product listings?

This actor reads only public search pages — the same pages any shopper sees without logging in — and takes only product facts: identifier, title, price, rating, review count, badges. No customer data, no reviewer identities, no seller names. Product prices and specifications are facts about goods on sale, not personal data. How you use them downstream is yours to assess.

FAQ

Why does it need a residential proxy? Amazon rejects datacenter addresses within a page or two and answers with a robot check served as HTTP 200. The actor treats that as a block and rotates, rather than reporting an empty run as a success.

Which currency are the prices in? The marketplace's own — amazon.de in EUR, amazon.co.uk in GBP. That is not automatic: Amazon quotes the currency of the country a request appears to come from, so the same amazon.de search returns INR from an Indian address and KRW from a Korean one. The run pins its session to the marketplace's country to prevent that, and if any row still arrives in another currency the run says so in its status rather than shipping a number that looks like euros and is not.

Why no seller name? A third-party Amazon seller is often a sole trader, which makes the shop name personal data under GDPR. We do not emit it, behind a flag or otherwise.

Why is price sometimes null? Some tiles show no price at all — variation parents, out-of-stock items, "see options". The row still carries the ASIN, title and rating so you can fetch the detail yourself.

Can I scrape specific ASINs instead of a keyword? Not in v1 — this starts from a keyword search. scrapeDetails will then open each matched product's page for you.

What does bestSellerRank mean? Amazon's position for that product inside bestSellerCategory, e.g. 2,190th in Computer & Zubehör. It is the field sellers watch, and it only arrives with scrapeDetails.

Why does reviewCount look like the rating? It should not — the reviews block prints the rating first and the count last, and the extractor takes the last number, with a test that fails if they are ever confused.

Troubleshooting

• "robot check instead of search results" — Amazon served its interstitial. The actor retries on a fresh route; re-run in a few minutes if a whole run fails.
• Fewer products than you expected — Amazon repeats the last page rather than ending cleanly, so the run stops as soon as a page adds nothing new.

If this saves you an afternoon, a rating on this page takes ten seconds and helps other buyers find the compliance-clean option.