The Best Amazon Reviews API in 2026 (We Tested Them)

We've pulled a lot of Amazon reviews. For competitor research, for sentiment dashboards, for agents that read the one-star reviews before a buyer commits. And every single time, the first thing someone asks is the same: what's the official Amazon reviews API?

There isn't one. Amazon's Product Advertising API gives you a star average and a review count, and not one word of review text. So "the best Amazon reviews API" is really a question about which scraper or vendor to trust, and how you want to pay for it.

Fair warning before we start: you're on the Monid blog, and one of the six options below is ours. We've listed it first because we think it's the best fit for most teams, but we'll be honest about the others, including the cases where they beat us. We rank each option on four things a review integration actually lives or dies on: does it return full text, how much setup and upkeep it costs you, whether it covers the fields and regions you need, and how the billing behaves when your usage is spiky. If you only need a star rating and a count, skip all of this and use the free official API. Everyone else, here are the real options.

There is no official Amazon reviews API. Everything that returns the actual text of a review is a scraper, so the only real questions are which one, and how you pay.

The short version

• The official API gives you ratings and counts, nothing else, and it's now retired as of May 2026.
• Building your own scraper is real engineering: proxies, CAPTCHAs, a login wall, and a parser that breaks the next time Amazon reshuffles the review page.
• Single-vendor APIs (Rainforest, Oxylabs, Axesso) are reliable but subscription-priced, so you commit before you know your volume.
• The pay-per-result route on Monid rents the same proven Apify actors per result, a fraction of a cent each, with no Amazon account, no proxy pool, and no monthly plan. Live prices at monid.ai/tools.

The best Amazon reviews APIs at a glance

• Monid (pay-per-result marketplace): best for most teams, and for agents
• Axesso: best for a straight ASIN-to-reviews feed
• Rainforest API: best for heavy, predictable daily volume
• Oxylabs: best for enterprise scraping with SLAs
• Amazon Product Advertising API: best when you only need ratings
• Build your own scraper: best when scraping is your actual product

Is there really no official Amazon reviews API?

Correct, and it trips up almost everyone. The Product Advertising API was built for affiliates who link to products, not for anyone who wants to read what buyers wrote. Its CustomerReviews field returns the star rating and the count, then hands you an iframe for the rest. No documented version of the official API has exposed full review text, and access depends on an Amazon Associates account with recent sales. So the moment your feature needs the text of a review, the official route is out, and you're choosing among scrapers.

If you're here because PA-API itself is gone, that's a separate move: Amazon retired it in May 2026, and we cover the migration in how to move off Amazon's PA-API. This post is about the reviews question specifically.

What makes a good Amazon reviews API

Before the list, here's what actually separates a useful option from a pile of JSON you have to babysit.

• Full review text, not a summary or a count. This is the line the official API can't cross.
• The verified-purchase flag. The most valuable field and the easiest to miss. It's your cheapest filter against planted or incentivized reviews.
• Source-side filtering by star, keyword, and reviewer type, so you pull the slice you need instead of downloading everything and throwing most away.
• Regional coverage. Amazon.com is one storefront. Selling in the UK, Germany, or Japan means you need amazon.co.uk, .de, and .co.jp too.
• A billing shape that matches your usage. Reviews work is bursty for most teams, so per-result pricing usually beats a subscription unless your volume is high and steady.

The fields a review record should return

A star rating tells you a product is a 4.3. The fields below tell you why the one-stars are one-stars, and which of them to trust. When you compare options, this is the payload to check for, because a feed missing the verified-purchase flag or the media links quietly costs you later.

Full body text and rating are the obvious ones. The verified-purchase flag is the field that separates signal from noise, since a review tied to a real order is worth more than an anonymous one. Helpful votes let you sort by what other buyers found useful. Attached media matters more than teams expect, because a photo of a cracked unit is a support signal a sentiment model reading text alone will miss. Variant association tells you whether the complaint was about the blue XL or the red small. And the aggregated rating distribution gives you the shape of sentiment in one number set, so you know whether the one-stars are a rounding error or a pattern.

The best Amazon reviews APIs in 2026

1. Monid (best for most teams)

Best for: teams and agents that want review text without a standing subscription or any scraping infrastructure.

Pros

• Rent proven Apify review actors per result, with the price shown before anything runs.
• Verified-purchase flag, star and keyword filters, and 20+ regional domains, none of which you have to build.
• Built-in redundancy: if one actor breaks or degrades, an alternative for the same job is already in the catalog, one discover away on the same wallet.
• One account and one wallet also reach social, search, and enrichment data, so reviews aren't a separate vendor.

Cons

• At genuinely enormous daily volume, a dedicated contract can beat per-result pricing on raw unit cost.
• It's a marketplace, so the underlying actor is run by a provider, not by Monid directly.

Monid is a pay-per-call data API marketplace: you discover an endpoint, inspect its schema and price for free, and only pay when you run it. For Amazon reviews that means pointing an actor at an ASIN and getting structured reviews back, billed per result (one ASIN query is a single result), with nothing committed up front. It's the option we'd reach for on any project that isn't pulling millions of reviews a day.

There's a quieter benefit that only shows up the day something breaks. With a single vendor, an outage or a broken actor is your outage, and you wait. On a marketplace, the same job usually has more than one endpoint: if the Axesso actor has a bad day, the web_wanderer extractor covers the same ASIN from the same wallet with one flag change, no new signup, no new invoice. Redundancy comes with the catalog instead of being something you architect.

Running it is one line once the CLI is installed, or a single instruction for an agent (set up https://monid.ai/SKILL.md):

monid keys add   # one time: paste your key from monid.ai
monid run -p apify -e /axesso_data/amazon-reviews-scraper \
  -i '{"input":[{"asin":"B0BSHF7WHW","domainCode":"com","maxPages":5,"sortBy":"recent"}]}' -w
# -> that ASIN's reviews with text, stars, dates, and verified-purchase
#    flags, billed per result, price shown by a free inspect first

Read more: pull an ASIN's whole review history in one call and score a product's reviews for sentiment in fifteen minutes.

2. Axesso (best for a clean ASIN-to-reviews feed)

Best for: a straightforward, well-structured feed of reviews by ASIN.

Pros

• Returns rating, title, full text, date, reviewer, verified-purchase flag, helpful votes, media, and the rating distribution.
• Filters by star, keyword, reviewer type, and media across regional domains.

Cons

• As a direct subscription it carries the usual commit-before-you-know-your-volume tradeoff.

Axesso is one of the more reliable Amazon data providers, and its review actor is the one we point to most often. You can buy it directly, or run the same axesso_data/amazon-reviews-scraper through Monid per result, which is how we'd start before committing to a plan.

3. Rainforest API (best for heavy, predictable volume)

Best for: teams pulling large Amazon volume every day on a fixed schedule.

Pros

• Built for scale and dependable at high throughput.
• Handles the anti-bot and infrastructure problems for you.

Cons

• Subscription pricing sized to your peak month, whether or not you use it.
• Another vendor, key, and invoice to manage alongside your other data sources.

Rainforest is a solid choice when Amazon data is a core, steady input rather than an occasional one. If your volume is genuinely high and consistent, a dedicated contract can come out cheaper per review than per-result pricing, and that's a real case worth pricing out.

4. Oxylabs (best for enterprise scraping with SLAs)

Best for: larger organizations that need scraping infrastructure with formal support.

Pros

• Enterprise-grade reliability, proxies, and support commitments.
• Broad scraping coverage beyond just Amazon.

Cons

• Priced and packaged for enterprise, so it's heavy for a small or bursty job.

Oxylabs sits at the infrastructure end of the market. If you have the volume and the need for SLAs, it's dependable. For a few thousand reviews now and then, it's more than the job requires.

5. Amazon Product Advertising API (best for ratings only)

Best for: showing a star rating and review count, nothing more.

Pros

• Free, official, stable, and Amazon's own.

Cons

• No review text, ever.
• Requires an Amazon Associates account with recent qualifying sales.

If your product genuinely only displays "4.6 stars, 12,431 ratings," this is the right answer and no scraper can match free-and-official. The instant you need the text behind those stars, though, you're back to the rest of this list.

6. Build your own scraper (best when scraping is your product)

Best for: teams whose core competency, and product, is scraping infrastructure.

Pros

• Total control and near-zero marginal cost per page.

Cons

• Proxies, CAPTCHAs, a login wall, and a parser you'll rewrite on Amazon's schedule.
• You build the verified-purchase flag, filtering, and regional support yourself, then keep them working.

Amazon is one of the hardest scrape targets on the web. Rolling your own can pencil out if this is what you do for a living. For everyone else, it's a pager that goes off during launches.

Other options worth knowing

For one-off jobs on arbitrary review pages (a Trustpilot page today, an Amazon listing tomorrow), Strale's product-reviews-extract bills a flat per-call fee and returns a ready-made sentiment and pros-and-cons summary. And if you need regional variants and aspect-level sentiment, web_wanderer/amazon-reviews-extractor reaches 20+ domains and tells you whether the one-star was about the blue XL or the red small. All three run on Monid: the two Amazon scrapers bill per result, Strale per call.

What teams actually build with Amazon review data

The API is a means to an end, and the end shapes which option fits. Four jobs come up again and again.

Competitor teardowns. Pull the one and two-star reviews for a rival's best-seller, cluster the complaints, and you have a feature gap list written by their own customers. This is bursty by nature, a burst before a launch and quiet after, which is exactly where per-result pricing wins.

Sentiment dashboards. Track a set of your own ASINs over time, score each new review, and watch the trend. Here the verified-purchase flag earns its keep, because you want the trend built on real buyers. See scoring an ASIN's reviews for sentiment for the mechanics.

Agent workflows. An agent that helps someone buy can read the recent one-stars and surface the real risks before the purchase, live, in the middle of a conversation. That only works if the reviews call is something the agent can make itself, which is the case on an MCP-native marketplace.

Support and product signal. Attached media and recurring phrases in reviews are a backlog nobody filed. Mining them turns a review feed into a product-and-support input, not just a marketing number.

What it costs, by shape

Cost is less about a single unit price and more about the shape of the bill, because that's what decides your total.

A subscription is a flat line. You pay your plan every month whether you pulled a million reviews or none, which is efficient only if your usage is high and steady enough to fill the plan. Pay-per-result is a line that tracks your actual work: a competitor teardown lands in pennies, a big sentiment backfill lands in single-digit dollars, and a month where you pull nothing costs nothing. Building your own has almost no marginal cost but a large fixed cost in engineering time and maintenance that never shows up on an invoice. For most teams, whose review work comes in bursts, the pay-per-result shape is cheaper across a full quarter even when its unit price looks higher on paper. Live per-endpoint prices are at monid.ai/tools.

So which should you use?

Start from the job, not the tool. Only need a rating and a count? PA-API, free, done. Pulling millions of reviews a day on a fixed schedule? Price out Rainforest or Oxylabs, because at that scale a flat plan can win. Anything in the wide middle, a few hundred to a few thousand reviews when a project calls for it, is where pay-per-result shines: real review text, the verified-purchase flag and filters already built, nothing committed up front, and the same balance ready the day you also need social or enrichment data.

The best Amazon reviews API is really just the one that gets you the text you need without a bill that keeps running when you don't. And you can always switch later. Run a free discover for "amazon product reviews" on Monid, pull a small batch on an ASIN you already know, and see if the data matches the listing.

Common questions

Is there an official Amazon reviews API? No. PA-API 5.0 returned only the star rating and review count, and the newer Creators API keeps that restriction, so review text always comes from a scraper.

What's the best Amazon reviews API in 2026? For review text with nothing committed up front, a pay-per-result Apify actor on Monid such as axesso_data/amazon-reviews-scraper fits most teams. Choose PA-API for ratings only, Rainforest or Oxylabs for heavy daily volume.

How much does it cost? On Monid, a fraction of a cent per result, shown before the call runs, with no subscription or minimum. Pulling one product's reviews is a single result, so even a batch of ASINs lands in pennies. Live prices are at monid.ai/tools.

Can I get reviews from Amazon outside the US? Yes. Both scraper endpoints read Amazon's regional storefronts, and web_wanderer/amazon-reviews-extractor covers 20+ domains, so UK, German, and Japanese listings come through the same call.

How do I filter out fake or incentivized reviews? Use the verified-purchase flag as your first gate, then weight by helpful votes. Neither is perfect, but a verified purchase with helpful votes is a far stronger signal than an anonymous five-star with none.

What if the scraper I'm using breaks? On a single-vendor subscription, you file a ticket and wait. On Monid, the catalog usually holds more than one endpoint for the same job, so you swap the provider flag (Axesso to web_wanderer, for example) and keep pulling from the same wallet while the broken one recovers.

Do I need proxies or an Amazon account? No, unless you build your own. Through Monid the proxies, retries, and layout fixes stay on the provider's side, and you bring one account and a pay-as-you-go balance.

Last updated July 2026.