Spot suspicious patterns in product reviews. Nothing ever uploads.

Paste a set of product reviews below — repeated phrases get grouped across the set, duplicates get flagged, and every review gets a transparent suspicion score with the exact signal that fired. Analyzed entirely on your device.

  • Nothing leaves your device
  • Up to 200 reviews at once
  • Repeated phrases grouped
  • Every signal explained

Analyze product reviews

Multi-line reviews? Pick "Blank line between reviews" and leave an empty line after each one.
No reviews yet. Everything stays on this device. Analysis is instant.

Privacy you can verify, not just trust

Most review checkers upload your pasted text to their servers first — and several keep it. This one can't: the analysis is plain JavaScript running in your browser. Your review text never leaves the device — the only third-party requests are the site's own analytics, which count page views — and it uploads nothing, ever. There is no file fetched after load for the analysis itself. Don't take our word for it:

New cross-origin requests during use: 0
  1. Press F12 to open developer tools, then the Network tab.
  2. Analyze any reviews above. Watch: apart from the site's analytics, nothing new is requested — no uploads, and no review text ever leaves your device.
  3. Extra credit: disconnect your internet after this page loads — analysis still works.

How to spot suspicious product reviews without uploading them

Between a third and two fifths of online reviews on major platforms are estimated to be inauthentic, and the best-known free checker, Fakespot, shut down in July 2025. The replacements that took its place mostly work the same way: paste a product URL, wait for a server to fetch the listing, and get back one opaque score. That works for Amazon listings — and it means handing the listing to someone else's server and trusting a number you cannot inspect. For reviews you already have in front of you, from any shop or platform, the safer approach is to analyze them where they already are: on your own device, with every signal visible.

This tool does exactly that. When you paste reviews here, your browser:

  1. Splits the paste into reviews — one per line, or separated by blank lines if you pick that setting — and reads an optional leading star rating ("5 stars", "5/5", "★★★★★") off each review, so a rating can be compared with the words that follow it.
  2. Groups repeated phrases — every four-word phrase that appears in two or more different reviews is listed with the reviews that carry it. The same wording showing up in several "independent" reviews is the strongest set-level signal there is.
  3. Scores each review against ten fixed signals — duplicates, incentivized wording, rating/text mismatch, generic praise, superlative overload, short-and-extreme, shouting, exclamation overload and detached wording — and shows the exact words that fired each one, so you can check the reasoning instead of trusting a number.

What makes this analyzer different

The pages that top the results for this search are all URL analyzers (ReviewMeta, RateBud, FakeFind, Null Fake) that fetch a product listing from their own servers, or single-review checkers (ReviewSense, Signed Reviews) that score one pasted review at a time, or a server-side scorer (Scamanot) that returns one opaque authenticity number for 3–10 pasted reviews. None of them is a dedicated free fully client-side tool that groups repeated phrases across a pasted set of reviews and scores each review with transparent, explainable signals — with no upload, no account and no limits. This page is the only one we found that combines phrase grouping across up to 200 pasted reviews, exact and near-duplicate detection, ten weighted signals with the firing words shown, an overall score with a plain-text report you can copy or download, and a live network-request counter so you can verify the privacy claim yourself. It runs entirely in your browser — there is no server to upload to.

When is this the right tool?

Whenever you are about to trust a set of reviews and something feels off: a product with dozens of five-star reviews that all sound the same, a listing where every review says "amazing product, highly recommend", or a seller whose reviews you want to sanity-check before buying. It also works the other way — sellers checking their own listings for patterns that make shoppers suspicious, and writers checking their own copy for repetitive phrasing. If you only need to check one review, paste it anyway: the per-review signals still run, though phrase grouping needs a set.

Honest limitations

A high score is not proof

The signals are text-only. They cannot verify a purchase, a reviewer identity, or posting times. A genuine review can score high — a short, enthusiastic review from a happy buyer looks like generic praise — and a careful fake can score low. Treat the score as a reason to look closer, never as a verdict.

Short reviews carry weak evidence

Several signals need a minimum word count, and a two-word review can only be scored on what it actually says. One short review proves almost nothing on its own; the set-level signals — repeated phrases and duplicates — are where the real evidence lives.

Fixed word lists

The superlative, generic-phrase and incentive lists are small and fixed. Slang, new phrasing and other languages can slip past them. The phrase-grouping and duplicate checks do not depend on the lists at all, so they keep working.

English lists only

The word lists are English. Repeated phrases, duplicates, exclamation overload and all-caps shouting work in any language, but generic praise, superlatives and incentivized wording are only detected in English text.

No product URL checking

This tool analyzes reviews you paste in. It does not fetch product pages, so it cannot check a listing by URL or see reviewer histories and posting dates. For an Amazon listing, a URL checker like ReviewMeta or RateBud covers that side — this tool covers the pasted-text side they do not.

Ratings are optional

The rating/text mismatch signal only fires when a review starts with a star rating ("5 stars", "5/5", "★★★★★"). Reviews pasted without a rating are scored on their wording alone, which is still most of the signal set.

Frequently asked questions

Do my reviews upload anywhere?

No. Your pasted reviews are analyzed entirely in your browser with plain JavaScript — your review text never leaves the device, and the only third-party requests are the site's own analytics, which count page views. You can verify it yourself with the network-request counter in the privacy section below the analyzer.

What counts as a repeated phrase?

A four-word phrase that appears in two or more different reviews — for example, "amazing product highly recommend" showing up in five reviews. Common filler is ignored: a phrase only counts when it contains at least one meaningful word, so "and it is very" does not qualify. The results table shows every repeated phrase, how many reviews carry it, and exactly which reviews those are.

Does a high score mean the review is fake?

No. A high score is a warning, not a verdict. The signals are text-only — they cannot verify a purchase, a reviewer identity, or posting times — so a genuine review can score high (a short, enthusiastic review from a happy buyer looks like generic praise) and a careful fake can score low. Use the score to decide which reviews deserve a closer look, not to declare any single review fake.

Why do I have to paste reviews — can't you check a product page by URL?

Checking a product page by URL means fetching that page from a server, which is exactly what this tool avoids: your analysis stays on your device, with no account and no limits. The URL-based checkers (ReviewMeta, RateBud, FakeFind) do that server-side work for Amazon listings. This tool is for the other case — reviews you already have in front of you, from any shop or platform, pasted in and analyzed locally.

What is an incentivized review?

A review written in exchange for something — a free product, a discount, or payment. The tool flags wording that gives it away, such as "in exchange for an honest review", "received this free", "free sample" or "sponsored". Incentivized reviews are not automatically fake — the reviewer may genuinely like the product — but they are not independent either, so they deserve a closer look.

How many reviews should I paste?

At least 5, and 10 or more is better. Repeated phrases and duplicates only show up when several reviews are compared, so a single review can only be scored on its own wording. The tool accepts up to 200 reviews at a time. If you only have one review to check, paste it anyway — the per-review signals (generic praise, superlative density, incentivized wording, rating mismatch) still run.

Does it work for reviews in other languages?

Only partly. The signal word lists are English, so generic praise, superlatives and incentivized wording are detected in English text only. But the language-independent signals still work in any language: repeated phrases, exact and near-duplicate reviews, exclamation overload and all-caps shouting are all computed from the text itself, not from English words.

What should I do with a suspicious result?

Read the reviews the tool flagged and judge for yourself: look for the repeated phrases in context, check whether the flagged reviews were posted close together, and see whether the same reviewer name appears more than once. If you are buying, treat a high overall score as a reason to look for other sources — the manufacturer's site, independent reviews, or a retailer with verified-purchase badges — before deciding. Copy or download the plain-text report from the results to keep a record.

Can I analyze reviews I wrote myself?

Yes — sellers and creators use it to check their own listings for review patterns that look suspicious to shoppers, and writers use it to spot repetitive phrasing in their own copy. Your text never leaves the device either way, and nothing is stored.