Amazon Reviews Scraper

Pull hundreds of real Amazon customer reviews per product as clean structured rows - across all five star ratings, deduped and newest-first. Drop in product IDs or paste product URLs, pick a marketplace, star filter and sort order, and export sortable, filterable rows ready for CSV, Excel or JSON. No HTML parsing, no anti-bot headaches. No subscription, pay only per review.

---

Why this Actor?

Data Forge pulls a deep, recent sample of real customer reviews - not just the handful of featured reviews most tools stop at.

What you get	Data Forge Amazon Reviews	Typical review actor
Reviews per product	Up to 300 recent	8-10 featured only
Star filter (five_star to one_star)	Yes, 8 options	No
Sentiment filter (positive / critical)	Yes	No
Sort order	Recent or most-helpful	Fixed
Marketplaces	8 (us, uk, de, fr, it, es, ca, jp)	1
Fields per review row	Title + text + rating + date + author + verified + photos	Rating + text
Pricing	Pay per review returned	Monthly subscription

What it does

Give it any mix of ASINs and/or Amazon product URLs and it pulls the customer reviews for each product (up to 800 each), with every field structured:

• ⭐ Star rating the reviewer gave (1-5)
• 📝 Title and 💬 full review text
• 👤 Reviewer name and 📅 date
• ✅ Verified Purchase flag and 🍃 Amazon Vine flag
• 👍 Helpful votes
• 🖼 Review photos (all attached images)
• 🌍 Country the review was written in

Narrow the result with the star filter (all stars, a single rating, or positive / critical sentiment) and choose most recent or most helpful ordering. Set a soft cap per product to keep runs quick and cheap, or pull up to 800 per product for a deep dataset. Flip on featured reviews to also grab the highlighted reviews shown on the product page.

Reviews are deduped across all five star ratings and both sort orders. Smaller products come back in full; very large products return the most-recent and most-helpful slice.

Input

{
  "products": ["B0FQFB8FMG", "https://www.amazon.com/dp/B09B8V1LZ3"],
  "country": "us",
  "starFilter": "all_stars",
  "sort": "recent",
  "maxReviewsPerAsin": 300,
  "includeTopReviewsFallback": false,
  "stopOnError": false
}

Field	What it does
Products (ASINs or URLs)	The products to pull reviews for - paste an ASIN (10-char code, e.g. B0FQFB8FMG) or a full product URL. Mix both freely.
Country	Amazon marketplace: us, uk, de, fr, it, es, ca, jp.
Star filter	all_stars (default) / five_star / four_star / three_star / two_star / one_star / positive / critical.
Sort	recent (default) or helpful.
Max reviews per product	Soft cap on reviews returned per product (1-800). A lower cap runs faster and cheaper - you only pay for reviews returned.
Also pull featured reviews	Additionally grab the highlighted reviews shown on the product page.
Stop on error	Abort the whole run on the first failed product, instead of skipping it.

Provide at least one product - an ASIN or a product URL - or the run fails fast with a clear message.

Input modes and example tasks

Four copy-paste recipes, matched 1:1 to the ready-made example tasks on this Actor's page. Each one runs on a single ASIN (B0DGHMNQ5Z) so you can try it in seconds, then swap in your own products.

1. Full sentiment dataset

Product and analytics teams: pull a broad recent set across all stars to score sentiment and surface recurring themes.

{
  "products": ["B0DGHMNQ5Z"],
  "maxReviewsPerAsin": 300
}

2. One-star complaint mining

Support and QA leads: isolate the harshest feedback to find defects, return drivers and refund triggers.

{
  "products": ["B0DGHMNQ5Z"],
  "starFilter": "one_star",
  "maxReviewsPerAsin": 300
}

3. Five-star social proof

Marketing and CRO teams: harvest glowing reviews for landing pages, ads and testimonial widgets.

{
  "products": ["B0DGHMNQ5Z"],
  "starFilter": "five_star",
  "maxReviewsPerAsin": 300
}

4. Most-helpful first

Researchers on a budget: lead with the reviews shoppers voted most helpful for the highest-signal sample per dollar.

{
  "products": ["B0DGHMNQ5Z"],
  "sort": "helpful",
  "maxReviewsPerAsin": 300
}

Output

One flat row per review, with the full raw review object under data. The dataset ships with a ready-made Reviews table tab, so the results read as a clean spreadsheet you can sort, filter and export to CSV / Excel / JSON in one click.

A real review row:

{
  "asin": "B0FQFB8FMG",
  "author": "Jordan M.",
  "rating": 5,
  "title": "Best earbuds I've owned",
  "text": "Battery life is excellent and the noise cancelling is...",
  "date": "Reviewed in the United States on March 14, 2026",
  "country": "us",
  "verified_purchase": true,
  "vine": false,
  "helpful_count": 12,
  "images": ["https://m.media-amazon.com/images/I/..."],
  "image_url": "https://m.media-amazon.com/images/I/...",
  "data": { "...": "full review object" }
}

Console columns on the Reviews tab include asin, rating, title, text, author, date, verified_purchase, vine, helpful_count and country. Error rows carry an error_code and are free - you only pay for real reviews.

A run-summary record (OUTPUT key) reports the review count, error count and estimated cost. Live per-event pricing is shown on this actor's Apify Store page.

Key features

• Hundreds of reviews per product - across all five star ratings, deduped and newest-first, up to 800 each.
• Every review field structured - rating, title, text, author, date, verified-purchase, Vine, helpful votes, photos, country.
• 8 marketplaces - us, uk, de, fr, it, es, ca, jp.
• Star + sentiment filters - all stars, any single rating, or positive / critical, sorted by recent or helpful.
• ASINs or URLs in one field - feed product IDs or paste product links; the ASIN is extracted for you.
• Fast cached responses - repeat pulls return quickly; competitors scrape live at 3+ seconds.
• You only pay for results - errors and your own misconfigurations are never charged.

FAQ

Is scraping Amazon reviews legal? The Actor reads only publicly visible review content - the same text, ratings and photos any shopper sees on the product page. It collects no private or personal data beyond what the review shows in public. You stay responsible for using the output in line with your local regulations and Amazon's terms; common uses are product research, sentiment analysis and competitor monitoring.

How fresh are the reviews? Reviews are pulled on every run and returned newest-first, so a July 2026 run surfaces reviews posted the same month - verified live, not served from an old archive. Repeat pulls of the same product may return a short-lived cache for speed; change any input to force a brand-new harvest.

How many reviews can I get per product? Set maxReviewsPerAsin anywhere up to 800. Because Amazon limits how many reviews are publicly retrievable, you receive a deep, recent sample - up to about 300 reviews per product - rather than every review ever posted. Smaller catalogs come back in full; large ones return the most-recent and most-helpful slice.

Can I schedule automatic runs? Yes. Use Apify Schedules to run the Actor hourly, daily or weekly and keep a rolling review dataset current. Combine it with webhooks or the dataset API to stream new reviews straight into your database, BI tool or spreadsheet.

Which marketplaces are covered? Eight: us, uk, de, fr, it, es, ca and jp. Pick one per run with the country field.

What exactly am I charged for? You pay per review returned - nothing else. Error rows and misconfigured inputs are free, so a blocked or invalid product never costs you a cent.

Related actors

• Amazon Products Scraper - full product details, pricing, ratings and specs for any ASIN or URL.
• Amazon Scraper - search, best-sellers and category crawling across all 8 Amazon marketplaces.
• Walmart Reviews Scraper - the same recent-review harvesting, built for Walmart product pages.

Support

Need help, higher limits, or a custom scraper built for your workflow? I personally support every customer - reach out on any channel and I answer fast.