Trustpilot Reviews Scraper: Reviews, Ratings, Company Replies & Filters

Get the Trustpilot reviews of any company from its domain: stars, title, text, date, language, the reviewer's country, the Verified badge, the company's reply, and the company's overall score and review count. Filter by stars, language, date, keyword, verified reviews and reviews with a company reply. No reviewer names, profiles or photos, and company replies are cleaned of people's names (greeting, signature, names inside a sentence). Built for reputation monitoring, customer feedback analysis, competitor benchmarks and market research.

From $0.44 per 1,000 reviews, plus $0.002 per company and a $0.005 start fee per run. A review that cannot be delivered is never charged. A 100-review run on one company costs $0.055 on the Free plan price and $0.051 at Gold.

Quick start

1. Add one or more companies: a domain (zalando.fr), a website address, or a Trustpilot page address.
2. Choose how many reviews per company and, if you want, filters.
3. Click Start and read the reviews in the Output tab.

It also runs from the Apify API, on a schedule, and connects to webhooks, Make, Zapier, n8n or Apify's MCP server. Run it every week on the same companies to follow the score and the complaints over time.

What it is used for

• Reputation monitoring: new low-star reviews of your brand, and whether the company replied.
• Competitor benchmark: score, volume and recurring complaints of rivals.
• Customer feedback analysis: feed reviews to a sentiment or topic model, by language and period.
• Market research: what customers of a whole sector praise or complain about.

What you get

Stars, title, review text	yes
Publication time, language	yes
Verified badge	yes
Country shown for the reviewer	when given
The company's reply and its time (people's names removed)	when there is one
Company name, overall score, total review count	on every review
Link to the review	yes
Reviewer name, profile, photo, staff names	never delivered

Inputs

• Companies: domains, website addresses or Trustpilot company page addresses. Up to 2,000 per run.
• Reviews per company: 1 to 1,000 (default 100).
• Order: newest first (default) or Trustpilot's relevance order.
• Stars: a minimum and a maximum (for example 1 to 2 for complaints).
• Languages: review language codes (en, fr, de...).
• Published after / before: dates like 2026-01-31.
• Verified only, only reviews the company replied to, keyword.
• Maximum reviews and Apify's Maximum cost per run: the run stops before going over either.

Output

One record per review. Main fields: companyDomain, companyName, companyRating, companyReviewCount, rating, title, text, language, publishedAt, verified, reviewerCountry, companyReply, companyReplyAt, hasCompanyReply, reviewId, reviewUrl, scrapedAt. A company that cannot be found, or that has no review matching your filters, comes as a free error or missing record with the reason. The run summary (OUTPUT) gives the count per company and why the run stopped.

Pricing

Pay per event:

Event	Price
Review delivered	$0.48 per 1,000 (Free, Bronze), $0.46 (Silver), $0.44 (Gold and above)
Company looked up	$0.002 per company (found or not found), charged once the lookup answers
Run start	$0.005 per run

The company fee pays for the lookup even when the company has no page or no review matching your filters, so a typo in a list never stops the others. Never charged: reviews that fail, error and missing records (only the company fee applies), repeats, notices. Example at Gold: 10 companies with 100 reviews each = 1,000 x $0.00044 + 10 x $0.002 + $0.005 = $0.465.

Limits

• Only the 200 newest reviews of a company can be read in one pass (or the 200 most relevant). For more than 200, the Actor takes the 200 newest, then reads each star rating further to complete them (up to 200 per star rating), so a company with a very long history is covered by "the 200 newest per star rating", not by every review ever written.
• Dates work within that reach. If the dates you ask are older than the newest reviews that can be read, the company comes as a free missing record that says so, or with a note when only part of your dates could be checked.
• Filters by stars, language, date, keyword, verified and company reply are applied by Trustpilot's own search where possible, and always checked on every review before it is delivered.
• Companies are found by domain. A name without a domain is not looked up.
• Trustpilot shows what it shows at that moment: reviews it removes or reorders are not recoverable.

FAQ

Why no names? They are personal data. Reviewer names are never delivered. Company replies are cleaned by a name recognition model (the main European languages): the name in a greeting and in a signature is removed ("Hi Daniel," becomes "Hi,"), and a name inside a sentence becomes [name]; the reviewer's own name is removed wherever it is left. Company names and teams ("N26 support") are kept; an email address in a reply can be replaced by [email]. The model is good, not perfect: a rare or oddly written name (for example a first name alone at the end of a line) can slip through, and an unusual word can be taken for a name. The review text itself is delivered as its author wrote it.

Is anything different on Apify's free plan? Live collection needs a paid Apify plan. On the free plan, the prefilled example (www.zalando.fr, 20 reviews, newest first, no filter) returns a stored sample, marked sample: true with its date. Any other input asks nothing and says so in the run. Only the run start is charged.

What if a run is restarted? It resumes where it was: reviews already delivered are not collected or charged again. A run you reboot or resurrect more than 3 times stops there.

What if the run's timeout comes first? The reviews already read are delivered and charged, the rest is not. Resurrect the run to finish it, or give the next run a longer timeout.

Does the Actor send anything to Quarra? Only when a run fails, runs out of time, or stops early because our review service is down: the Actor, its build, the error type, the line of our code where it happened, and the memory and time the run was given, so we can fix it fast. Never your companies, your results, your account or the run ID.