nvNeuralVerge
E-commerce & Marketplaces

AliExpress Product Reviews API and MCP — Buyer Reviews, Auto-Translated

Send an AliExpress product ID or URL and this source returns the product's buyer reviews as data, one record per review, with a filter for review type and an option to machine-translate the text.

You send
An AliExpress product ID or URL, with optional filter and translate
You get back
Individual buyer reviews with rating, text, author and images
Coverage
AliExpress products with a public reviews section.
Price
1 point per returned result, up to 250 per request — $1 per 1,000
Billed per review actually returned, capped by max_items. A product with no reviews costs nothing.
Comparison

Ways to get this data

Three approaches teams use today, and where each one runs into trouble.

ApproachWhat it takesCostTrade-off
Reading reviews by handOpen the listing, filter, scroll, translate manuallyFree, and slowFine for one product, useless across a catalog — and most reviews aren't in your language
Building a scraperWrite it, host it, and add translation yourselfEngineering timeA maintenance commitment for something that isn't your product
NeuralVergeOne request per product$1 per 1,000, misses not billedReturns what's published — reviews are self-selected, not a representative sample
Response

What you get back

Every field in the example response below.

FieldTypeDescription
total_itemsnumberHow many reviews were returned in this batch (capped by max_items).
reviewsarrayIndividual review records, when the product has any that match the filter. Each carries id, text, rating, buyer_name, created_at_text (the date as AliExpress shows it, e.g. "04 Oct 2026"), sku_text, images, up_vote_count, is_anonymous and labels, plus buyer_country, buyer_gender, buyer_image, created_at, sku_id, logistics, down_vote_count, additional_text, additional_at_text and additional_images, which can be null or empty. In the example, buyer_country, buyer_gender, buyer_image, created_at, sku_id, logistics, down_vote_count and the additional_* fields are null or empty.
debugnullNull in the example, where reviews were returned. When no reviews come back it holds a diagnostic snippet (page title, URL and a text excerpt) to help explain the empty result.
Response

Example response

A real response, as returned.

{
  "session_id": "11111111-1111-1111-1111-111111111111",
  "kind": "aliexpress_product_reviews",
  "product": "1005012791281455",
  "max_items": 2,
  "language": null,
  "currency": null,
  "filter": "all",
  "translate": true,
  "human": "# AliExpress product reviews\n\n- **Product:** 1005012791281455\n- **Max items:** 2\n- **Records returned:** 1\n- **Proxy traffic:** 2.15 MB",
  "machine": {
    "total_items": 1,
    "reviews": [
      {
        "@type": "AliexpressProductReview",
        "id": "0",
        "text": "There was a slight delay in delivery, but the store was very transparent and everything worked out well. The mouse arrived in perfect condition, is very beautiful, and works well.",
        "rating": 5,
        "buyer_name": "M***z",
        "buyer_country": null,
        "buyer_gender": null,
        "buyer_image": null,
        "created_at": null,
        "created_at_text": "04 Oct 2026",
        "sku_id": null,
        "sku_text": "Color:Pink",
        "logistics": null,
        "images": [
          "http://ae-pic-a1.aliexpress-media.com/kf/UT8NVOIXRhXXXagOFbX2.jpg_220x220.jpg_.avif"
        ],
        "up_vote_count": 0,
        "down_vote_count": null,
        "is_anonymous": false,
        "additional_text": null,
        "additional_at_text": null,
        "additional_images": [],
        "labels": []
      }
    ],
    "proxy_traffic_bytes": 2254950,
    "debug": null
  },
  "total_points": 1
}
Integration

How to run it

The same lookup works from the app, the API, or as a tool an agent can call mid-task.

product accepts either an AliExpress product ID or URL. filter narrows to all/image/additional/local, and translate (default true) machine-translates review text — set it to false if you want the original language.

curl -X POST https://api.neuralverge.ai/functions/v1/run-aliexpress-product-reviews \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "product": "1005012791281455", "max_items": 10, "filter": "image" }'
Use cases

What teams use it for

Where this field shows up in an actual workflow.

Sourcing due diligence

Check recent buyer reviews before sourcing a product to resell.

Quality monitoring

Watch review volume and rating for a product you already sell.

Visual proof screening

Filter to image reviews to see what buyers actually received.

Cross-language research

Read non-English reviews in translation without a manual lookup per review.

Coverage & limits

Coverage, freshness and limits

What this source covers, how fresh it is, and where it stops.

Coverage

AliExpress products with a public reviews section.

Freshness

Read at request time.

Handling

This source returns publicly visible review content, stamped with the product ID or URL it was read from and a session id you can trace.

What it will not do
  • —Reviews are self-selected and skew towards strong opinions in both directions.
  • —Machine translation can miss nuance, particularly for idioms or product-specific terms.
  • —New or low-volume listings return few or no reviews.
FAQ

Questions, answered

Yes — translate defaults to true, since a large share of AliExpress reviews aren't written in English. Set it to false to get the original text instead.

Related

Related sources

Try aliexpress product reviews on your own data

One request format across every source in the catalog.

Get started