Best Amazon scrapers in 2026

An Amazon scraper is a tool that requests Amazon pages for you and returns the product data on them (title, price, rating, seller, availability) in a format you can store or analyze. Some return raw HTML, others return clean JSON.

The one thing to know before you buy: the same ASIN shows different prices, stock and delivery promises depending on where the request appears to come from. A scraper that cannot pin a delivery location gives you data that looks fine and is quietly inconsistent. Everything else (price, speed, free tier) comes second.

All eight tools in our ranking list a dedicated Amazon scraper, endpoint, actor or template on their own site, so none of them is just a generic fetcher pointed at amazon.com.

If you can write code and need volume, an Amazon scraper API is usually the better fit. If you need a one-off sheet of products, an Amazon web scraper you set up with clicks is faster. The four approaches below trade control for effort.

An Amazon price scraper is a scheduled job that fetches the same product pages at a fixed interval and stores each price with a timestamp and a location. The scheduling and the storage matter as much as the scraping.

Keep the list of ASINs in a table, run the job at the interval your decision needs (hourly for repricing, daily for reporting), and write every result as a new row instead of overwriting the last one. Store the marketplace and ZIP code in the same row, otherwise you will not be able to explain a price drop that was only a change of location.

Browse AI's scheduled monitoring suits small watchlists. For a few thousand ASINs, a structured endpoint from ScraperAPI, ScrapingBee, Scrape.do or Decodo is easier to maintain, and you can bill-check it per successful page. For the cost side, our free scraping APIs ranking shows which tools let you test before paying.

An Amazon product scraper returns the fields on a product page, while an Amazon review scraper returns customer reviews, which are often a separate request. Do not assume one endpoint gives you both.

Our notes confirm Amazon coverage for each vendor but do not break out review endpoints one by one, so check each tool's sample output before you commit. The fields people usually need are:

You set the Amazon domain (amazon.com, amazon.de, amazon.co.uk) in the URL and the delivery location through the tool's geo or ZIP parameter, then keep both identical between runs.

Check the docs for two things: whether the endpoint accepts a ZIP code or only a country, and whether the location is applied to the page or only to the proxy IP. Scrape.do lists geo-targeting on every plan and Decodo lets you choose proxies per request, but whether a specific Amazon endpoint takes a ZIP is something to confirm in each tool's own documentation.

A cheap test: request the same ASIN with two different ZIP codes and compare price and delivery date. If the output is identical, the targeting is not doing what you think. Residential IPs help with location realism; see our residential vs datacenter proxies guide for the trade-off.

Because Amazon detects automated traffic by IP reputation, request pattern and browser fingerprint, and answers suspicious requests with a captcha page or a stripped-down response. A plain script from one IP hits this quickly.

Scraping APIs exist to absorb this: they rotate IPs, retry failed requests and render JavaScript when needed. The general scraping APIs in our ranking list JavaScript rendering and residential IPs, but each one meters them differently. ScraperAPI and ScrapingBee use credits that cost more for rendering and harder modes, and Crawlbase bills per successful request by domain complexity.

That is why the cost per successful page, not the headline price, is the number to track. Log three counts during your test: pages returned correctly, pages returned as captcha or empty, and credits spent. If a tool charges for failures, the third number will be higher than you expect.

There is no single answer: collecting public pages is lawful in many places, but Amazon's terms of service, copyright on descriptions and images, and privacy rules for reviewer names still apply.

Amazon also runs official programs for sellers and affiliates, with their own conditions. Courts and platform rules differ by country. This is a description, not legal advice, so ask a lawyer before you build a business on scraped data.

Start with the cheapest tool that has a free tier, test it on 100 real product URLs, and divide the bill by the number of successful, correct results.

In our data, Decodo starts at $19/mo, SerpApi at $25/mo, Scrape.do at $29/mo with 1,000 free credits and no card, ScraperAPI and ScrapingBee at $49/mo, and Crawlbase at $149/mo. Apify and Browse AI have no entry price verified yet, so measure their real cost on the same sample.

Check three things in the results: credits used per product page (JS rendering often multiplies it), whether the location you asked for was applied, and how many pages came back empty or as a captcha. Then compare in the scraping API ranking and run the same 100 URLs on a second tool before you scale.