Amazon Product Scraper API - Prices, ASINs & Sellers
DeprecatedPricing
from $1.00 / 1,000 product scrapeds
Amazon Product Scraper API - Prices, ASINs & Sellers
DeprecatedAmazon scraper that extracts product data without using the Amazon API. Scrape prices, ASINs, ratings, review insights, bestseller ranks, sellers and variants from any search, category or product URL across 24 Amazon marketplaces. Download as JSON, CSV, Excel or XML.
Pricing
from $1.00 / 1,000 product scrapeds
Rating
0.0
(0)
Developer
Divya Raj
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
3 days ago
Last modified
Categories
Features
Share
Amazon Product Scraper API — prices, ASINs, reviews and sellers
Scrape Amazon product data without using the Amazon API. Give this Amazon scraper any search, category or product URL and it returns prices, ASINs, ratings, review insights, bestseller ranks, sellers and variants — across 24 Amazon marketplaces — as JSON, CSV, Excel or XML.
No login. No cookies. No Amazon Product Advertising API key. Call it from Python, JavaScript or plain HTTP through the Apify API, or send the results straight into n8n, Make, Zapier or Google Sheets.
55 fields per product, including list prices, star breakdowns, Amazon's own AI review summary, bestseller ranks, sellers with their registered business details, every variant with its own price, A+ content, and delivery dates for any postcode.
What you can use it for
| Use case | What to pull |
|---|---|
| Price monitoring & repricing | price, listPrice, shippingPrice, offers, per-variant prices |
| Competitor & market research | A whole category or search term, with categoryPageData.productPosition and isSponsored |
| Product research for sellers | bestsellerRanks, monthlyPurchaseVolume, reviewsCount, starsBreakdown |
| Review intelligence | aiReviewsSummary — Amazon's own sentiment per aspect, with the quoted reviews behind it |
| Supplier and brand audits | seller with business name, registered address and per-period ratings |
| Catalogue & feed building | attributes, productOverview, features, galleryThumbnails, highResolutionImages |
| Availability & logistics | inStock, inStockText, delivery, fastestDelivery for any postcode |
How to use it
- Paste one or more Amazon URLs into Start URLs.
- Set Max results per start URL.
- Run it. Each product becomes one dataset row.
Three kinds of URL work:
| URL shape | Example |
|---|---|
| Search | https://www.amazon.com/s?k=keyboard |
| Category | https://www.amazon.com/s?i=electronics&bbn=16225007011&rh=n%3A16225007011 |
| Product | https://www.amazon.com/dp/B09X7MPX8L |
Amazon marketplaces supported
.com · .co.uk · .de · .fr · .it · .es · .nl · .se · .pl · .com.be · .ie · .com.tr · .ae · .sa · .eg · .in · .sg · .co.jp · .com.au · .ca · .com.mx · .com.br · .cl · .co.za
Every one of those 24 was checked on a live run: the right currency, the marketplace's own language, its country in loadedCountryCode, and a localised delivery address. Each page is pinned to its marketplace's language and currency by cookie, so a German product reads in German and euros whatever country the proxy exits from, and the proxy country follows each URL automatically.
amazon.cn is not supported — Amazon closed its China marketplace in 2019 and the domain no longer carries a product catalogue.
Output
One row per product. Abbreviated — the full record carries 55 fields:
{"title": "AULA F75 Pro Wireless Mechanical Keyboard","url": "https://www.amazon.com/dp/B0D14N2QZF","asin": "B0D14N2QZF","brand": "AULA","price": { "value": 68.99, "currency": "$" },"listPrice": { "value": 79.99, "currency": "$" },"inStock": true,"inStockText": "In Stock Only 2 left in stock - order soon.","stars": 4.4,"starsBreakdown": { "5star": 0.75, "4star": 0.12, "3star": 0.05, "2star": 0.02, "1star": 0.06 },"reviewsCount": 111682,"monthlyPurchaseVolume": "5K+ bought in past month","breadCrumbs": "Video Games > PC > Accessories > Gaming Keyboards","bestsellerRanks": [{ "rank": 56, "category": "Video Games", "url": "https://www.amazon.com/gp/bestsellers/videogames/ref=pd_zg_ts_videogames" }],"galleryThumbnails": ["https://m.media-amazon.com/images/I/31kI-vgQwYL._AC_US.jpg"],"highResolutionImages": ["https://m.media-amazon.com/images/I/61a14rKCeCL._AC_SL1500_.jpg"],"videosCount": 16,"variantAsins": ["B0D83DMWFC", "B0DS7JXKYD"],"variantDetails": [{ "asin": "B0D83DMWFC", "name": "F75 Pro-Gradient Purple", "price": { "value": 69.99, "currency": "$" } }],"variantAttributes": [{ "key": "Color", "value": "F75 Pro-Black" }],"attributes": [{ "key": "Connectivity Technology", "value": "Bluetooth, USB" }],"productOverview": [{ "key": "Brand", "value": "AULA" }],"aiReviewsSummary": {"text": "Customers find the keyboard to be of good quality…","keywords": [{"name": "Quality","sentiment": "positive","text": "Customers find it well built.","customersMentionedCount": { "total": 412, "positive": 353, "negative": 59 },"partialReviews": [{ "text": "great keyboard except for the gaps", "highlightedPart": "great keyboard" }]}]},"seller": {"id": "A9TV4QL8DFMM5","name": "AULA MALL","businessName": "Shenzhen Suoai Electronics Technology co.,Ltd","address": ["518103", "CN"],"ratingLifetime": { "starsOutOf5": 4.9, "ratingCount": 1626, "percentage5Star": 93 }},"offers": [{ "position": 1, "isPinnedOffer": true, "condition": "New", "price": { "value": 68.99, "currency": "$" }, "shipsFrom": "Amazon" }],"aPlusContent": { "title": "From the manufacturer", "rawText": "…", "modules": [] },"brandStory": { "title": "From the brand", "image": { "name": "BS", "url": "…" }, "items": [] },"delivery": "Tuesday, November 17","fastestDelivery": "Tomorrow, November 13","locationText": "Deliver to New York 10001","loadedCountryCode": "US","categoryPageData": { "pageNumber": 1, "productPosition": 3, "isSponsored": false },"input": "https://www.amazon.com/s?k=keyboard"}
Fields Amazon does not render for a product come back as null, never missing, so the shape stays stable across marketplaces and categories. Book, grocery and apparel fields (author, bookDescription, condition, support, manufacturerAttributes, priceRange, importantInformation, sustainabilityFeatures) fill in on the categories that have them.
Input
| Field | Type | Default | Purpose |
|---|---|---|---|
categoryOrProductUrls | array | — | Required. Search, category or product URLs. |
maxItemsPerStartUrl | integer | 100 | Result cap, per start URL. |
maxSearchPagesPerStartUrl | integer | 9999 | Page cap, per start URL. |
scrapeProductDetails | boolean | true | Off = grid data only. Far faster and far cheaper. |
maxOffers | integer | 0 | Offers per product. See the note below. |
scrapeSellers | boolean | false | Merchant profiles: business name, registered address, storefront, ratings per period. |
maxProductVariantsAsSeparateResults | integer | 0 | Emit variants as their own rows, each a full record. |
scrapeProductVariantPrices | boolean | false | Price every variant. |
language | string | — | Amazon UI language; overrides the marketplace default. |
proxyCountry | string | auto | Follows the URL's marketplace unless you set it. |
countryCode / zipCode | string | — | Delivery location. Prices, stock and delivery dates follow it. |
locationDeliverableRoutes | array | all | Which page types the delivery location applies to. |
proxyConfiguration | object | Apify proxy | Standard Apify proxy settings. |
useCaptchaSolver | boolean | false | Accepted, deliberately not implemented — see below. |
Delivery location
Amazon prices and stocks differently by address. Set zipCode (and countryCode) and the scraper performs the same address change the site's own location widget does, once per session:
zipCode: "10001" → locationText: "Deliver to New York 10001"fastestDelivery: "Today 7 PM - 11 PM"
What maxOffers returns today
The featured offer — the buy box — is built from the product page itself, so maxOffers: 1 costs no extra request. Offers beyond it live in a panel Amazon now loads from an endpoint that answers 404 to a direct request; nine URL shapes were tried with the page's own headers and session cookies. Until that is solved, a product with several sellers returns the featured offer only. This is stated here rather than discovered after a run.
Why useCaptchaSolver does nothing
Amazon's own documentation for this option notes it works only on .com, and that solving a captcha makes Amazon stop rendering attributes, manufacturerAttributes and bestsellerRanks. Requests here start on datacenter proxies and only a retry goes through residential, which brought the captcha rate on a 50-product crawl to zero. The option would buy nothing and cost three fields.
Speed and cost
Measured on 50-product crawls of amazon.com:
| Proxy strategy | Time | Captchas | Platform cost / 1,000 products |
|---|---|---|---|
| Datacenter only | 110 s | 29% of requests | $0.36 |
| Residential only | 48 s | none | $3.24 |
| Datacenter first, residential on retry — the default | 55 s | none | $0.68 |
The cheapest lever is scrapeProductDetails: false. A search page yields 16–48 products in one request, so the grid fields — title, ASIN, price, rating, review count, thumbnail, position — cost a fraction of a request each. Leaving details on costs one request per product.
scrapeSellers, scrapeProductVariantPrices and maxProductVariantsAsSeparateResults each add requests per product, which is why all three are off by default.
What you pay for
Charging is per unit of work, so you pay for what a run actually did and nothing is bundled into a price you cannot opt out of:
| You are charged for | When |
|---|---|
| A product scraped | One full 55-field record from a detail page |
| A product from a listing | One grid row, when scrapeProductDetails is off — a fraction of the above |
| A seller profile | Only with scrapeSellers, and once per seller however many of their products appear |
| A variant priced | Only with scrapeProductVariantPrices, once per variant |
| A page rendered in a browser | Only on the marketplaces that need it, and the session is reused so later runs rarely do |
Error rows are never charged. Current prices are on the Actor's page above.
Error rows
An input that cannot be scraped produces an error row rather than vanishing, so you can reconcile inputs against outputs. Normal rows never carry an error field.
{"error": "product_not_found","errorDescription": "Loaded a 404 page. The product does not exist.","input": "https://www.amazon.com/dp/B0XXXXXXXX","url": "https://www.amazon.com/dp/B0XXXXXXXX"}
error | Meaning |
|---|---|
invalid_url | A start URL was malformed or not recognised |
invalid_input | Bad configuration; the run is terminated too |
product_not_found | Product URL returned 404, or could not be retrieved |
shortened_url_invalid | An Amazon short link could not be resolved |
bestsellers_category_not_found | Bestsellers category page not found |
no_relevant_reviews_found | Reviews exist but none match the filters |
no_results_found | The input returned no results, or could not be retrieved |
A page that could not be retrieved keeps the code matching its page type and explains itself in errorDescription, so it is never mistaken for a verdict about the product.
Accuracy notes
- Price can be missing when no seller ships to the delivery country; setting
proxyCountryusually fixes it. - Amazon prices differ by US state and by proxy location. Two runs minutes apart from different exits will differ on
price,delivery,fastestDeliveryandlocationText. That is Amazon, not the scraper. - Amazon caps some search terms at 7 pages. To go deeper, combine a category with the term rather than searching alone — instead of
Diamond rings, open theJewelrycategory and add the term. - Result counts fluctuate run to run; Amazon applies its own internal caps. Run a small test scrape to size your use case.
Integrations
Results land in an Apify dataset, so they can be pulled as JSON, CSV, Excel or XML over the API, or pushed straight into n8n, Make, Zapier, Airbyte, Google Sheets, Google Drive, Slack, GitHub and anything else that speaks webhooks.
Is scraping Amazon legal?
Scraping publicly available data such as product titles, descriptions, prices and ratings is legal. This Actor collects no personal data and requires no login. If you plan to republish what you collect, check the copyright position for your jurisdiction.
FAQ
Is there a free Amazon scraper? Apify's free tier covers a trial run, so you can test this one before paying anything. Beyond that you pay per product scraped — the price is on the Actor page above, and error rows are never charged.
Can I use it from Python?
Yes. Start a run with the apify-client package for Python or JavaScript, or with a plain HTTP POST to the Apify API, then read the dataset as JSON or CSV. Nothing here is tied to a language.
Is this the Amazon Product Advertising API? No, and it needs no key for it. It reads the same public pages a browser does, so it returns fields the official API does not expose — bestseller ranks, the AI review summary, per-variant prices and seller business details.
Does it work on amazon.co.uk, amazon.de, amazon.in?
Yes, and on 21 more. Each marketplace returns its own language, currency and delivery dates. The list is above.
Can it scrape a whole category?
Yes. Give it the category URL and raise maxItemsPerStartUrl. Amazon's own 7-page cap applies to bare search terms, not to categories.
Can I get prices for a specific city or postcode?
Yes — set zipCode. Prices, availability and delivery dates are then the ones that address sees.
Does it need a browser?
Only two marketplaces need one. amazon.fr and amazon.in answer a plain request with a JavaScript challenge, so those are rendered, and the session that establishes is reused afterwards — later runs reach both over plain HTTP and no browser starts. Everything else is plain HTTP, which is why it is cheap.
How fresh is the data? It is read from Amazon at the moment of the run. Nothing is cached.
Can I scrape reviews?
Review insights are included — aiReviewsSummary carries Amazon's own per-aspect sentiment, mention counts and the quoted reviews behind each aspect, plus starsBreakdown and reviewsLink. Full review-by-review scraping is not in scope.
Development
npm installnpm test
Runs on CheerioCrawler — plain HTTP plus gzip. Amazon's product markup is server-rendered, so for all but two marketplaces a headless browser would add cost without adding data. A Playwright pass handles the rest, and only the requests that failed without it.