Amazon Bestsellers Scraper — Category Ranks, Prices & ASINs

Amazon category-ranking research for product teams

Use this Amazon Best Sellers scraper to collect an observed category list for a selected marketplace. Merchandising teams can create a dated category worksheet, researchers can identify products for deeper inspection, and analysts can compare saved ranking observations. A category position is not a measurement of unit sales, profit or advertising performance.

Workflow	Keep in the export	Avoid assuming
Review a category shortlist	categoryUrl, domain, categoryName, rank	That every top-100 item was accessible
Join a product research table	ASIN and productUrl where exposed	That a title alone uniquely identifies a product
Inspect displayed offers	price, rating and reviewsCount strings	That locale-formatted strings are normalized numeric data
Compare dated observations	Category, marketplace, rank and your own collection timestamp	That a rank move proves a change in sales
Compare selected markets	Separate categories and domain context	That ranks in different markets describe the same population

For a first request, use one explicit Best Sellers category URL and maxPages=1. The input supports up to two page requests per category. Actual distinct rows can be lower, and the current Actor has no maxItems parameter, resume token or built-in historical ledger.

MCP and AI agent guide

Retrieve the observed Amazon Best Sellers list for a category and marketplace. This is category ranking, not keyword search, verified sales or a full product catalogue.

Connect an authenticated Apify MCP client to:

https://mcp.apify.com/?tools=apivault_labs~amazon-bestsellers-scraper

Use your own Apify token in the client Authorization header; never paste it into Actor inputs, prompts or public examples. Call tools/list and use the current schema for apivault_labs--amazon-bestsellers-scraper. The input description summarizes the first-run/output contract; this README contains the full workflow.

Minimal Actor input:

{
  "categoryUrls": [
    "https://www.amazon.com/Best-Sellers-Books/zgbs/books"
  ],
  "maxPages": 1,
  "maxConcurrency": 1
}

For a direct Actor MCP call, add waitSecs: 45 to the tool arguments if the discovered schema supports it. waitSecs is a tool wait, not an Actor input, run timeout or spending limit. If the response is still RUNNING, keep its runId and poll get-actor-run; do not launch a duplicate paid run. After a network timeout, reconcile the existing run before repeating the call.

Use returned default storage IDs. Read get-dataset-items with a small limit (start with 5) and selected fields; use get-key-value-store-record only for the records supported below. Fetch a small page first, then paginate the Dataset instead of filling model context with the whole export. The wrapper may return a preview before the run finishes. Confirm terminal status, actual rows and result validity before reporting success. A missing record is not a zero value.

Agent decisions and output contract

• Provide category Best Sellers URLs in categoryUrls, or a supported category slug with domain. Use an explicit URL for the first call. maxPages is 1 or 2; it is a page request limit, not a guarantee of 50/100 complete ranked products.
• This tool does not expose maxItems, keyword search, city filters, a continuation token or a history baseline. Do not invent those inputs. Repeated runs create new observations; compare saved observations yourself if tracking changes.
• One Dataset row is an observed product/category result. Preserve categoryUrl, domain, categoryName, rank, asin and productUrl when joining data. A product can occur in more than one category; rank belongs to its category and observation.
• price, rating and reviewsCount are source display strings. Parse locale/currency explicitly when needed; missing price or rank is unknown, not zero. Extraction switches can omit fields.
• The Actor does not write SUMMARY or ERRORS records. Read the Dataset and run log/status. Source refusal or an unavailable category can yield SUCCEEDED with an empty Dataset: this does not prove there are no bestsellers. Do not silently retry a paid run or claim completeness from an empty response.
• At current rates, useful Dataset results cost $2/1K, plus the configured start fee. Separately paid platform usage is active. Check the current Pricing tab for applicable rates. Keep the first request to one category/page and confirm costs before expanding.

Use the smallest request that answers the user's question. Costs and source availability are separate from MCP discovery: seeing a schema does not prove a successful collection or image generation. Full instructions may be shortened by discovery summaries; open this Store README when more detail is needed.

⚡ Use this Actor in n8n — no code

Building no-code workflows in n8n? Install our community node and drop this Actor straight onto your canvas:

📦 n8n-nodes-apivault-amazon-bestsellers

Setup: in n8n go to Settings → Community Nodes → Install, paste n8n-nodes-apivault-amazon-bestsellers and confirm → add the new node to your workflow → paste your Apify API token → fill in the input and run.

Scrape accessible Amazon Best Sellers category pages — request a bounded category-ranking sample with rank, price, rating, reviews count, ASIN and image. Perfect for dropshipping research, trend spotting, and product discovery.

✨ Key Features

• 🏆 Top-ranked products by category (rank 1 to 100)
• 💵 Price, ⭐ rating, reviews count
• 🔢 ASIN auto-extracted (use with our Amazon Product Scraper for deeper data)
• 🌍 Supports 12 marketplace choices in the input; current access and field coverage vary
• 📂 Any category or subcategory
• 🧲 Auto-deduplication
• 🔒 No Amazon API key, no login

Input

Request up to two electronics category pages

{
  "categoryUrls": [
    "https://www.amazon.com/Best-Sellers-Electronics/zgbs/electronics"
  ],
  "maxPages": 2
}

Multi-category research

{
  "categoryUrls": [
    "https://www.amazon.com/Best-Sellers-Beauty/zgbs/beauty",
    "https://www.amazon.com/Best-Sellers-Kitchen/zgbs/kitchen",
    "toys"
  ]
}

International (Germany bestsellers)

{
  "categoryUrls": ["electronics"],
  "domain": "amazon.de"
}

Input Parameters

Field	Type	Required	Description
categoryUrls	string[]	✅	Full bestsellers URLs or category slugs (electronics, toys)
domain	string	❌	Amazon marketplace when only slug provided (default: amazon.com)
maxPages	int	❌	Request 1 or 2 source pages. Actual distinct rows and rank coverage can be lower.
extractRank / extractTitle / extractPrice / extractRating / extractReviewsCount / extractAsin / extractProductUrl / extractImage	bool	❌	Toggle fields (all default true)
extractCategoryName	bool	❌	Include category name on each row (default: true)
maxConcurrency	int	❌	Parallel page scrapes (default: 2)
timeout	int	❌	Timeout per page (default: 180)

Output example — illustration

The following is a fictional shape example, not a live price/rank observation. Fields depend on the source and extraction switches.

{
  "success": true,
  "categoryUrl": "https://www.amazon.com/Best-Sellers-Electronics/zgbs/electronics",
  "categoryName": "electronics",
  "domain": "amazon.com",
  "rank": 1,
  "title": "Apple AirPods Pro (2nd Generation)",
  "price": "$189.99",
  "rating": "4.7 out of 5 stars",
  "reviewsCount": "123,456",
  "asin": "B0BDHWDR12",
  "productUrl": "https://www.amazon.com/dp/B0BDHWDR12",
  "imageUrl": "https://m.media-amazon.com/images/I/..."
}

Use Cases

• Dropshipping research — find trending products to sell
• Niche validation — check which categories are hot
• Affiliate marketing — build top-10 lists automatically
• Competitor intelligence — track where competitors' products rank
• Price monitoring — watch bestseller prices over time
• Product launch research — see what's dominating before launch

Pricing

• $0.002 per item ($2 per 1,000 items)
• Pay only for successfully extracted items
• Empty or duplicate entries do not add result events; startup and platform usage can still apply

How it works

Builds Amazon bestsellers URL from your category, requests up to two category pages, collects publicly available category observations, deduplicates by ASIN per category.

Typical page windows per category (coverage is not guaranteed):

• 1 page → up to 50
• 2 pages → up to 100

Popular Category Examples

Category	URL slug
Top Overall	(leave empty or use "all")
Electronics	electronics
Beauty	beauty
Kitchen	kitchen
Toys & Games	toys-and-games
Books	books
Home & Kitchen	home-garden
Clothing	fashion
Health	hpc
Pet Supplies	pet-supplies

Tips

• Combine with Amazon Product Scraper: first pull top ASINs here, then scrape detailed product data with our Amazon Product Scraper
• Track trends: run this daily and diff results — products moving up are gaining momentum
• International arbitrage: compare bestsellers across amazon.com / amazon.co.uk / amazon.de
• Keep maxConcurrency: 2 for reliability

Repeatable Amazon bestseller research inputs

Rank belongs to the requested category and marketplace. Keep those fields when joining ASINs across categories. Full category URLs select their own domain; the domain option is used for bare category slugs. Schema support for a marketplace does not establish a successful fresh collection from every country.

One-page books category observation

{
  "categoryUrls": [
    "https://www.amazon.com/Best-Sellers-Books/zgbs/books"
  ],
  "maxPages": 1,
  "maxConcurrency": 1,
  "timeout": 120
}

A bounded first request can still return a page of products rather than a single row. maxPages limits source page requests, not delivered items or fees. Set the platform run charge cap when a monetary ceiling is required.

Use a UK category slug explicitly

{
  "categoryUrls": [
    "books"
  ],
  "domain": "amazon.co.uk",
  "maxPages": 1,
  "maxConcurrency": 1,
  "timeout": 120
}

This requests a category in the UK context. Prices/review counts remain source display strings; do not combine them with US strings as if currencies and formatting were identical. This is a schema-valid example, not a claim that all locales were tested in this release.

Keep a smaller set of result fields

{
  "categoryUrls": [
    "https://www.amazon.com/Best-Sellers-Books/zgbs/books"
  ],
  "maxPages": 1,
  "maxConcurrency": 1,
  "extractImage": false,
  "extractRating": false,
  "extractReviewsCount": false,
  "extractCategoryName": true
}

This omits three optional fields while preserving the other defaults, including identity and rank where available. Output selection does not reduce the per-row price or guarantee that a missing ASIN can be recovered.

Make a useful category snapshot

Export the Dataset as JSON or CSV, add your application's collection timestamp, and retain the exact category URL and input settings. Rank, title, displayed price, rating and review-count observations can be incomplete. Parse currency and number formatting in a separate validated step; an absent price is unknown, not a free product. A missing rank should not be treated as rank zero.

Deduplication uses category and ASIN, with a content fallback when identity is unavailable. It is not a global catalogue identity service. The same ASIN appearing in two categories can legitimately have two rows with different rank context. If extractAsin is disabled or the source omits it, review identity before merging observations rather than assuming the fallback is a verified ASIN.

For trend work, keep comparable dated observations in your own database. This Actor does not automatically diff runs, persist rank history, estimate sales velocity or trigger price alerts. Category membership, page availability and source presentation can change between observations. A product disappearing from a bounded response does not prove that it left the category or stopped selling.

Costs, empty output and automation questions

Is every top-100 product guaranteed? No. maxPages=2 requests the supported second page, but source access and extraction determine actual rows. Use the output and run logs to state the observed window. Do not extrapolate a partial category page into a complete catalogue or market-size estimate.

Can I cap results with maxItems? That input is not exposed. Use one category/page for a small test and configure the run's event-charge limit. Source work may happen before delivery is capped, so platform usage is still separate.

What is the current fee? Each delivered Dataset row is $0.002, plus the $0.00005 start event at the standard memory size. Ten delivered rows create $0.02005 in event fees before separately billed compute/proxy/storage. Empty/duplicate rows do not create extra result events; a failed or empty run is not necessarily free of startup and resource cost. Caller-paid platform usage is active as of this documentation check.

Why can an empty run say SUCCEEDED? The current runtime can finish with an empty Dataset after unavailable source data. It does not write SUMMARY or ERRORS records. Inspect actual rows, logs and run status; do not invent records or interpret an empty response as proof that the category contains no products. A known source refusal needs an access change before another attempt.

Are the sample output prices live? The output illustration above is fictional documentation, not a current Amazon quote or a benchmark. This release's bounded QA records its actual build, counts and limitations separately. A successful sample does not establish complete coverage for other categories.

Does the community node work in n8n Cloud? npm publication establishes a community package, not verified-node status or Cloud availability. Use the integration links below for the matched package and source. An automation should start one run, poll its returned ID and read its returned Dataset instead of repeatedly creating runs while waiting.

n8n and developer integrations

Use this Actor in your automation or application with the existing community integrations below. An Apify account and API token are required; Actor and platform charges follow the current pricing.

• n8n community node: n8n-nodes-apivault-amazon-bestsellers. Install package name n8n-nodes-apivault-amazon-bestsellers in a self-hosted n8n instance that permits community packages.
• Python client and examples: GitHub.
• n8n source and usage: GitHub.

These are community packages. Publication on npm does not establish n8n verification or availability in n8n Cloud. Review the linked package README for supported inputs and setup.