Amazon Product Scraper — GPSR, Variants, Offers & A+ Content

What does Amazon Product Scraper do?

Amazon Product Scraper extracts full, typed product data from Amazon across 18 marketplaces — by ASIN, product URL, or keyword search — including pricing, ratings, Best Sellers Rank, buy box, variations, frequently-bought-together bundles, Subscribe & Save pricing, A+ content, and free monthly sales and revenue estimates for niche research. It is the only Amazon product scraper that also delivers EU GPSR compliance data (manufacturer, EU responsible person, safety warnings) and an audit-proof provenance block (URL, HTTP status, SHA-256 of the page) on every single item — at no extra charge. Whether you need price intelligence, GPSR compliance monitoring for EU sellers, or keyword-based niche research with sales estimates, you get clean, analysis-ready JSON with per-field coverage status that never invents a value.

The actor runs on the Apify platform, so you get API access, scheduling, integrations, proxy rotation, and run monitoring out of the box — no infrastructure to manage. Pricing is pay per product with no platform-usage passthrough, and errors are never charged. Try it on the Apify free tier in a couple of clicks.

What makes this Amazon scraper different?

Most Amazon scrapers stop at title, price, and BSR. This one was built for data you can defend — in an audit, in a compliance report, or in a pricing model.

Capability	Amazon Product Scraper	Typical Amazon scraper
GPSR / EU compliance data (manufacturer, EU responsible person, safety warnings)	✅ included free	❌ not available
Provenance / capture block (URL, HTTP status, SHA-256 page hash)	✅ on every item	❌ not available
Per-field coverage status (obtained / absent / unknown)	✅ on every item	❌ silent nulls
Marketplaces	18	1–3
Variant price matrix (every variation child with typed price)	✅ opt-in add-on	⚠️ partial or missing
Seller profiles with legal imprint	✅ opt-in add-on	❌ not available
A+ content structure (premium flags, brand story)	✅ included	❌ not available
Sales & revenue estimates	✅ model-based, included free	❌ not available
Keyword search mode with auto-pagination	✅ built in	⚠️ rare
Pricing	Pay per product ($0.004), no platform-usage passthrough, errors never charged	Often per-result plus compute/usage fees

GPSR compliance monitoring for EU sellers

Since the EU General Product Safety Regulation (GPSR) took full effect in December 2024, Amazon requires manufacturer and EU responsible-person information on EU listings — and has been deactivating non-compliant listings since 2025, reportedly within 24 hours of detection. For EU sellers and the agencies managing their catalogs, GPSR data is no longer a one-off check but a recurring monitoring task.

This actor extracts the GPSR block live from the product page's compliance panel on EU marketplaces:

• Manufacturer — name and postal address as displayed
• EU responsible person(s) — name and postal address as displayed
• Product safety images and warnings — the safety pictograms and documents Amazon renders in the compliance panel

In a live benchmark of 10 ASINs on Amazon.de (August 2026), the actor delivered GPSR data for 9 of 10 products — on the tenth, Amazon verifiably does not render the compliance panel — while two major competing product-data actors returned no GPSR fields at all. GPSR fields are part of the standard product event: no add-on, no extra charge. Combined with the per-field coverage status and the provenance capture block (URL, HTTP status, SHA-256 of the page), every run doubles as audit-proof evidence of what a listing showed at a point in time — schedule it daily or weekly to watch your whole catalog, or your competitors'.

How much does it cost to scrape Amazon products?

The actor uses pay-per-event pricing — you pay only for delivered results, and there is no platform-usage passthrough: unlike some competing Amazon scrapers, we do not bill you for compute units or proxy traffic on top. The event prices below are the whole cost (exact prices are always shown on the actor's Pricing tab in Apify Console):

• product — $0.004 per product, charged once per successfully scraped product item. Includes GPSR data, provenance, FBT bundles, Subscribe & Save, A+ content, and sales estimates.
• offers (opt-in) — $0.0025 per ASIN, charged only when the offer list was requested and actually delivered. If it can't be delivered, the reason is recorded in offers_error and there is no charge.
• variant-prices (opt-in) — $0.0015 per ASIN, charged only when the complete variant price matrix was delivered. A partial matrix is attached with variant_prices_incomplete but never charged, as are ASINs without variations.
• seller-details (opt-in) — $0.0025 per ASIN, charged only when at least one seller profile was delivered. Undelivered profiles are reported in seller_details_partial / seller_details_error.
• apify-actor-start — a one-time $0.01 per run.

Errors are never charged. Invalid inputs, not-found pages, blocked requests, and transport failures produce transparency items in the dataset at zero cost.

Cost by volume (product event only, + $0.01 per run)

Products per run	Base price	With all three add-ons*
100 products	$0.41	$1.06
1,000 products	$4.01	$10.51
10,000 products	$40.01	$105.01
100,000 products	$400.01	$1,050.01

* All add-ons enabled and delivered on every product ($0.004 + $0.0025 + $0.0015 + $0.0025 = $0.0105 per ASIN). Add-ons are charged per ASIN only when delivered, so real-world cost with add-ons is typically lower. The Apify free tier is enough to try the actor on a handful of ASINs before you scale up.

Who is this Amazon scraper for?

Audience	What they use it for
EU compliance officers & agencies	Recurring GPSR monitoring of client catalogs — manufacturer, EU responsible person, safety warnings — with SHA-256 provenance as audit evidence.
FBA sellers & product researchers	Keyword search mode + sales/revenue estimates for niche research: rank search results, spot sponsored placements, shortlist candidate ASINs, then deep-scrape them.
Brand teams	Listing and content QA at scale: A+ content structure, brand story, bullets, images, and "Customers say" aspects with verbatim Amazon sentiment counters.
Pricing analysts	Locale-correct typed prices, list price, discount, price per unit, buy box, full offer lists, and the complete variant price matrix across 18 marketplaces.
AI / ML teams	Clean, typed, provenance-stamped product JSON for training data, retrieval pipelines, and agents — with honest null / "unknown" semantics instead of hallucinated gaps.

How to scrape Amazon product data with Amazon Product Scraper

1. Open the actor in Apify Store and click Try for free.
2. On the Input tab, provide ASINs (e.g. B079VP6DH5) and/or Amazon product URLs (/dp/... or /gp/product/...) — or switch to keyword mode with keywords like ["garlic press", "bamboo organizer"].
3. Pick the Marketplace — domain, locale, currency, and price parsing all follow this setting.
4. Optionally enable Include offers for the offer list, Include variant prices for the full variant price matrix, Include seller details for seller profiles, and Expand variants to scrape every variant child with full data.
5. Click Start, then download your results from the Output or Storage tab — or pull them via the Apify API.

Using the Apify API (JavaScript)

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: '<YOUR_APIFY_TOKEN>' });

const run = await client.actor('CyprusAPI/amazon-product').call({
    keywords: ['garlic press'],
    maxResultsPerKeyword: 50,
    marketplace: 'com',
    includeOffers: true,
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items);

Using the Apify API (Python)

from apify_client import ApifyClient

client = ApifyClient('<YOUR_APIFY_TOKEN>')

run = client.actor('CyprusAPI/amazon-product').call(run_input={
    'keywords': ['garlic press'],
    'maxResultsPerKeyword': 50,
    'marketplace': 'com',
    'includeOffers': True,
})

for item in client.dataset(run['defaultDatasetId']).iterate_items():
    print(item)

Input

The actor accepts the following parameters (see the Input tab for the full form). At least one of asins, startUrls, or keywords is required — keyword mode searches the run's marketplace and scrapes the found ASINs as full product items, charged as normal product events.

Field	Type	Description
asins	array	Amazon ASINs (10 characters) to scrape. One dataset item per ASIN.
startUrls	array	Amazon product URLs; the ASIN is extracted from each URL. Provide asins and/or startUrls.
keywords	array	Keyword mode: search terms to run on the marketplace's search. Each found ASIN is scraped as a full item and tagged with search_keyword, search_position, and sponsored.
maxResultsPerKeyword	integer	Cap on unique ASINs scraped per keyword, with auto-pagination. Default 50, min 1, max 500.
marketplace	string	One of 18 marketplaces: com, de, co.uk, fr, it, es, ca, com.au, co.jp, in, nl, se, pl, be, mx, br, ae, sa. Default com.
includeOffers	boolean	Also fetch the full offer list per ASIN. Charged as a separate offers event, only when delivered. Default false.
includeVariantPrices	boolean	Also resolve the full variant price matrix per ASIN (variant_prices). Charged as a separate variant-prices event, only when the complete matrix was delivered. Default false.
includeSellerDetails	boolean	Also fetch seller profiles (seller_details) for every offer seller per ASIN — feedback ratings per period, positive feedback, legal imprint. Charged as a separate seller-details event. Default false.
expandVariants	boolean	Enqueue every child ASIN of a variation parent as its own full product item. Default false.
maxItems	integer	Global upper bound of product items per run, across all keywords and other inputs, including expanded variant children. Default 10.
language	string	Optional Amazon language code (e.g. en_US, de_DE) appended to every fetched page.
proxyConfiguration	object	Proxy settings. Residential proxies are strongly recommended and preselected.
proxiesByMarketplace	object	Optional geo-pinned custom proxies per marketplace — tried first, ahead of the Apify proxy tiers.

Output

One dataset item per scraped product. Flat overview fields (title, price, rating, BSR, availability, image) sit next to full nested groups: identity, core, pricing, buybox, ratings, rank, demand, content, variations, customer_aspects, attributes, compliance, plus coverage and capture. Items found via keyword mode carry top-level search_keyword, search_position (1-indexed within that keyword's results, across pages), and sponsored (null for items from asins/startUrls or variant expansion). Every item's demand group includes the free estimates bought_in_past_month, estimated_monthly_sales, estimated_monthly_revenue, and estimate_method. You can download the dataset in various formats such as JSON, HTML, CSV, or Excel.

{
  "asin": "B079VP6DH5",
  "url": "https://www.amazon.com/dp/B079VP6DH5",
  "marketplace": "com",
  "title": "Bounty Paper Towels Quick Size, White, 16 Family Rolls = 40 Regular Rolls",
  "brand": "Bounty",
  "price_current": 43.49,
  "price_list": null,
  "discount_percent": null,
  "currency": "USD",
  "rating_value": 4.8,
  "rating_count": 102341,
  "bsr": 12,
  "availability": "In Stock",
  "in_stock": true,
  "image": "https://m.media-amazon.com/images/I/714Rq4k05UL._AC_SL1000_.jpg",
  "search_keyword": "paper towels",
  "search_position": 7,
  "sponsored": false,
  "demand": {
    "bought_in_past_month": "10K+",
    "estimated_monthly_sales": 10000,
    "estimated_monthly_revenue": "$434,900",
    "estimate_method": "amazon_badge"
  },
  "content": {
    "bullets": ["16 Family Rolls = 40 Regular Rolls ...", "..."],
    "aplus": { "present": true, "is_premium": false, "module_types": ["standard-image", "..."], "brand_story_present": true }
  },
  "variations": {
    "parent_asin": "B07XYZ1234",
    "dimensions": ["Size"],
    "children": [{ "asin": "B079VP6DH5", "dimension_values": { "Size": "16 Family Rolls" } }],
    "child_count": 4
  },
  "frequentlyBoughtTogether": {
    "items": [{ "asin": "B0CB97R25C", "position": 1, "is_current_product": true, "price": 12.99, "price_string": "$12.99" }],
    "item_count": 3,
    "total_price": 38.84,
    "total_price_string": "$38.84",
    "currency": "USD"
  },
  "subscribeAndSave": {
    "price": 8.07,
    "price_string": "$8.07",
    "currency": "USD",
    "tiered_price": 7.22,
    "tiered_price_string": "$7.22 with 28 percent savings",
    "tiered_savings_percent": 28
  },
  "coverage": { "pricing": "obtained", "compliance": "absent", "customer_aspects": "obtained" },
  "capture": {
    "captured_at_utc": "2026-01-15T09:30:00+00:00",
    "url": "https://www.amazon.com/dp/B079VP6DH5",
    "http_status": 200,
    "was_authenticated": false,
    "content_hash": "sha256:1a2b3c..."
  },
  "scrapedAt": "2026-01-15T09:30:05+00:00"
}

The demand estimates work like this: bought_in_past_month is Amazon's badge text verbatim (e.g. "10K+"); when a badge exists, estimated_monthly_sales is its bucket floor ("10K+" → 10000) with estimate_method: "amazon_badge"; otherwise a BSR-model heuristic is used with estimate_method: "bsr_model". estimated_monthly_revenue is monthly sales × current price as a formatted string in the product's currency with marketplace separators (e.g. "$434,900", "3.745 €"). All four fields are null when no estimate is possible.

On EU marketplaces the compliance block additionally carries GPSR data (manufacturer, EU responsible person, safety warnings) extracted live from the product page. The frequentlyBoughtTogether and subscribeAndSave blocks are part of the standard product event (no add-on charge); they are currently verified live on amazon.com and amazon.de, using language-stable selectors across the other marketplaces.

Fields follow a three-state discipline: a value means observed, null means verifiably absent, and "unknown" means not observable anonymously — the coverage block reports this per field group. Failed ASINs appear as transparency items with error and errorDescription and are never charged.

Data table

Field	Description
asin, url, marketplace	Product identity and canonical URL
title, brand	Product title and brand (plus byline, brand store, breadcrumb in core)
price_current, price_list, discount_percent, price_per_unit, currency	Typed numeric prices, locale-aware parsing
rating_value, rating_count	Star rating and review count, with scope (marketplace/worldwide) and histogram in ratings
bsr	Best Sellers Rank (all nodes and badges in rank)
demand	Free estimates: bought_in_past_month (verbatim badge), estimated_monthly_sales, estimated_monthly_revenue, estimate_method (amazon_badge / bsr_model) — model-based, see Known limitations
search_keyword, search_position, sponsored	Search context for items found via keyword mode (null for ASIN/URL/variant-expansion items)
availability, in_stock	Availability message and stock flag
buybox	Featured offer, seller, offer count, fulfillment
offers	Full offer list: seller name/ID, price, condition, shipping, delivery, ships from (with includeOffers)
variant_prices	Variant price matrix: every variation child with typed price and dimension values (with includeVariantPrices)
seller_details	Seller profiles of the offer sellers: ratings per period, positive feedback, legal imprint (with includeSellerDetails)
variations	Parent ASIN, dimensions, all children with dimension values
frequentlyBoughtTogether	"Frequently bought together" bundle: component ASINs with typed prices, bundle total, currency — null when Amazon doesn't render the module
subscribeAndSave	Subscribe & Save price and tiered-unlock price (verbatim from Amazon's tiered label, e.g. "$7.22 with 28 percent savings") — null when not offered
content	Bullets, description, images, video, A+ structure with premium flags and brand story
customer_aspects	"Customers say" aspects with verbatim Amazon sentiment counters
attributes	Merged specifications from overview, product-information tables, and detail bullets
compliance	GPSR block, legal notices (warranty, seller), product documents (PDFs)
coverage	Per-field-group status: obtained / absent / unknown
capture	Provenance: timestamp, URL, HTTP status, user agent, SHA-256 content hash

Known limitations

We'd rather you know these up front:

• Sales and revenue estimates are model-based heuristics, not measured sales. When Amazon shows a "bought in past month" badge, the estimate is the badge's bucket floor (estimate_method: "amazon_badge"); otherwise a BSR-curve fallback is used (estimate_method: "bsr_model"). Treat estimated_monthly_sales and estimated_monthly_revenue as an order-of-magnitude ranking signal for comparing products — not as validated accuracy figures. We deliberately publish no error percentages, because we don't have ground-truth sales data to validate against.
• No review text. This actor extracts product-page data (including rating histograms and "Customers say" aspects), not individual customer reviews. For review text, use our sibling Amazon Reviews Scraper, which goes beyond Amazon's 100-review display limit.
• Browser-based rendering is slower than pure-HTTP scrapers. The actor renders pages in a real browser, which costs some speed but reliably handles Amazon's bot-wall layouts and JS-rendered modules (GPSR panel, A+ content, FBT) that plain HTTP scrapers often miss.
• Variant pricing is an opt-in add-on. The full variant price matrix requires includeVariantPrices (separate variant-prices event) because resolving every child's price takes extra requests; ASINs without variations are never charged for it.
• One marketplace per run. All ASINs in a run are fetched from the same marketplace; split multi-marketplace jobs into separate runs.

Tips and advanced options

• Use maxItems to cap run size and cost — it's the global cap across all keywords and inputs, and matters especially with expandVariants, where a single parent can enqueue dozens of children (deduplicated and capped).
• Proxy fallback chain. Each ASIN is attempted in tiers: your own proxiesByMarketplace pool first, then Apify datacenter proxies, then Apify residential. Retries only happen on bot-detection/transport failures (not_found is final). If your proxyConfiguration already uses residential groups, the datacenter hop is skipped — and Apify residential availability (incl. on the free plan) depends on your account plan.
• Match proxies to the marketplace. Residential proxies are preselected; with proxiesByMarketplace you can pin geo-located exit IPs per marketplace. A mismatched exit IP makes Amazon serve wrong currencies or bot-gates.
• Force a language with language (e.g. en_US) if you need consistent copy across locales — by default the marketplace's native language is used.
• Automate it. Schedule recurring runs, chain the actor in Apify integrations, or call it from your stack via the Apify API.

FAQ

Will Amazon block me? Do I need proxies?

Amazon actively rate-limits and bot-gates scraping traffic, which is why residential proxies are preselected and each ASIN is retried through a tiered fallback chain (your proxies → datacenter → residential) on bot-detection. Blocked requests surface as transparency items with error / errorDescription and are never charged.

Can I bring my own proxies?

Yes. Add your own endpoints via proxyConfiguration.proxyUrls, or pin geo-located proxies per marketplace with proxiesByMarketplace (e.g. German exits for amazon.de) — your pool is always tried first.

How fresh is the Best Sellers Rank?

BSR is scraped live at run time from the product page — it's exactly the rank Amazon displayed at the capture moment recorded in capture.captured_at_utc. Amazon refreshes BSR roughly hourly, so scheduled runs give you a true time series.

How accurate are the sales and revenue estimates?

Honestly: they're a model-based ranking signal, not measured sales. Badge-floor-derived when Amazon shows "bought in past month", BSR-model-based otherwise — the estimate_method field (amazon_badge vs bsr_model) tells you which signal each value rests on. Use them to rank and compare products within a niche, not as precise revenue figures — see Known limitations above.

Is scraping Amazon product data legal?

Web scraping is legal when you extract publicly available data that is not behind a login. This actor scrapes only anonymous, public product pages (was_authenticated is always false in the provenance block). Note that personal data is protected by regulations such as GDPR in the EU — you should have a legitimate reason to scrape personal data. If you are unsure, consult a lawyer. Also review Amazon's Terms of Service before scraping. Apify is not liable for how you use the extracted data.

Why did some fields come back as null or "unknown"?

That is the honest-data design, not a bug: null means the field is verifiably absent on the page, "unknown" means it could not be observed anonymously. The coverage block tells you which case applies per field group — no invented values.

Something isn't working — where do I report it?

Open the Issues tab on the actor's page in Apify Console and describe the problem, ideally with the failing ASIN, marketplace, and run ID. Error items in the dataset (error / errorDescription) already tell you what went wrong for individual ASINs.

Can you build a custom Amazon data solution?

Yes — if you need different fields, additional marketplaces, or a fully managed pipeline, reach out via the Issues tab and ask about a custom solution.