TikTok Shop Reviews Scraper

Collect public US TikTok Shop customer reviews for product research, customer-feedback analysis and variant comparisons. Submit product links or numeric IDs and get one structured row per review, ready for JSON, CSV or Excel export.

Pricing: $1 per 1,000 reviews

$1 per 1,000 successfully delivered reviews on the Free plan. Bronze: $0.95; Silver: $0.90; Gold, Platinum and Diamond: $0.85 per 1,000. Failed products, duplicate reviews and reviews excluded by your filters are not billed as reviews. There is a separate $0.0025 Actor-start event (one per GB of memory, minimum one). The start fee applies even when products are unavailable. There are no third-party API-key or subscription requirements.

At default memory, 10 reviews cost $0.0125 on the Free plan or $0.012 on Bronze. Set Apify's maximum total charge to control your result budget (minimum $0.0035). Standard Apify pricing and account terms still apply; consult the Pricing tab for current rates. This Actor uses no billable residential proxy by default.

Quick start

{
  "startUrls": ["https://shop.tiktok.com/us/pdp/1731578642912612516"],
  "maxReviews": 10,
  "sortBy": "recommended"
}

Product links from www.tiktok.com/shop/pdp/… and shop.tiktok.com/us/pdp/…, bare product IDs, and { "url": "…" } objects are accepted. Duplicate product inputs are skipped.

Review fields

Each review retains its original review_id and product_id as strings, numeric star rating, review text, raw publication timestamp, a UTC reviewDate, publicly displayed reviewer name and avatar, image URLs, SKU ID and specification, country, and explicit verified-purchase and incentivized-review flags. Product context includes the title, overall rating and accessible review total where supplied.

Missing purchase flags are null, not false. Public reviewer names may be masked by TikTok; this Actor does not unmask them. Empty-text ratings remain valid reviews. Media URLs may expire. Photo coverage is supported; video attachment coverage is not claimed.

Collection options

• maxReviews: maximum matching reviews per product, from 1 to 10,000.
• sortBy: source-recommended order or newest-first.
• starRating: 0 for all reviews, or an exact rating from 1 to 5.
• mediaOnly: keep only reviews containing photo URLs.
• verifiedPurchaseOnly: require TikTok's explicit verified flag.
• since: optional inclusive UTC date, formatted YYYY-MM-DD.
• maxPagesPerProduct: bound the pages requested when filters are restrictive.
• maxRetries: bounded retry count for temporary source errors.

Use a page cap appropriate to your budget. A filtered product may require many requests to find matching reviews, or have none. Limits are per product; your Apify maximum total charge caps the complete run.

Reports and partial results

REPORT records per-product delivery counts, page counts, total reviews, source failures, duplicate/filtered/invalid counts and whether the spending limit stopped collection. ERRORS lists sanitized product-level failures. Delivered records are saved incrementally. Resume checkpoints preserve delivered IDs to reduce duplicate results after migration.

If every product fails, the run fails explicitly. Successful products can retain their data when another product fails. An accessible product's review total is not a promise that every review can be retrieved. The page cap, filters, source restrictions and your spending limit can all reduce coverage.

Scope and limitations

Only US TikTok Shop products are supported. The Actor is independent and is not endorsed by TikTok. It uses the public storefront; it does not access orders, private messages or seller accounts. No customer TikTok login is requested. Explicit login or CAPTCHA challenges are reported, not solved. Regional restrictions, removed products and source changes can cause partial or failed runs.

Use the data responsibly and respect applicable platform terms and privacy requirements. This is an extraction tool, not a sentiment-scoring service; it does not invent reviews, infer demographics or guarantee sales insights.

API and support

Run technicaldost/tiktok-shop-reviews-plus with the Apify API, client SDK or integrations, then read the run's default dataset and REPORT record:

import os
from apify_client import ApifyClient

client = ApifyClient(os.environ["APIFY_TOKEN"])
run = client.actor("technicaldost/tiktok-shop-reviews-plus").call(
    run_input={
        "startUrls": ["1731578642912612516"],
        "maxReviews": 30,
        "sortBy": "newest"
    },
    max_total_charge_usd=0.05
)
reviews = client.dataset(run["defaultDatasetId"]).list_items().items
report = client.key_value_store(run["defaultKeyValueStoreId"]).get_record("REPORT")
print(len(reviews))

Open an issue on the Actor's Issues tab with the product URL, input and run ID. Never include cookies, access tokens or passwords in a public issue.

How collection works

A single browser is reused across products. Each product has an isolated public session so TikTok's normal page scripts can initialize; reviews then come from the storefront's own JSON review feed, in pages of 10. Image, font and video downloads are blocked to reduce bandwidth. Returned records are normalized and filtered locally, with IDs deduplicated before saving and charging. No paid third-party review service is wrapped, and no seller or customer credentials are used.

Startup typically takes longer than a plain HTTP scraper. Speed varies with source response times; no throughput or success-rate guarantee is claimed.

Frequently asked questions

Can it scrape any TikTok Shop market? No. Only US product detail pages are currently supported.

Does it guarantee every review? No. The public feed may restrict access or change. The report distinguishes limits, page caps, exhaustion and source failures.

Are the reviews generated or sentiment-scored? No. These are source records. Use your own analysis tools for sentiment.

Why are some reviewer names hidden? TikTok masks public identities. We preserve the displayed name; we do not reveal private account information.

What happens if the budget runs out? Collection stops with the affordable delivered rows retained and spendingLimitReached in the report.