Capterra Reviews Scraper: All Reviews, Software Advice, GetApp
Pricing
from $0.60 / 1,000 reviews
Capterra Reviews Scraper: All Reviews, Software Advice, GetApp
Scrape every Capterra review of a software product, also via Software Advice and GetApp: ratings, pros, cons, reviewer industry and company size, date. Gets past Capterra's 100-page limit. Export to CSV, Excel or JSON. New-reviews-only monitoring and webhook. No login, no proxy.
Pricing
from $0.60 / 1,000 reviews
Rating
0.0
(0)
Developer
Alexis Morain
Maintained by CommunityActor stats
0
Bookmarked
1
Total users
1
Monthly active users
4 days ago
Last modified
Categories
Features
Share
Capterra Reviews Scraper
This Actor exports the reviews of any software product listed on Capterra, Software Advice and GetApp. Paste a product URL or type a product name and you get one row per review: ratings and sub-ratings, pros, cons, the reviewer's industry and company size, how long they used the product, whether the review was incentivized, the vendor's reply and the date. A summary row per product is added.
It works as a Capterra reviews API: no login, no browser, no proxy, plain HTTP requests from Apify servers.
Its main point is the full history. Capterra lists 25 reviews per page and stops at page 100, so a scraper that pages through a product reaches about 2,500 reviews and nothing tells it the rest exists. Software Advice, where this Actor reads, stops at 10,000 per listing. This Actor splits big products into slices that the site filters itself, reads every slice and removes duplicates. On 6 November 2026 it returned 23,893 of the 24,225 reviews of Slack, where paging stops at 10,000. Products under 10,000 reviews come out in full.
Pricing is pay per event: $1.00 per 1,000 reviews on the Free plan, down to $0.60 on Gold, plus $0.005 per run and $0.003 per product summary row (optional). A run of 20 reviews costs about three cents.
A new reviews only mode is built for monitoring: each scheduled run returns, and charges, only reviews it has not delivered before, and a webhook can notify your own tool.
What does this Capterra reviews scraper do?
You give it products: a Capterra URL such as https://www.capterra.com/p/135003/Slack/ (regional sites such as capterra.co.uk work too), a Software Advice URL such as https://www.softwareadvice.com/collaboration/slack-profile/, a GetApp URL such as https://www.getapp.com/collaboration-software/a/slack/, or a plain product name such as HubSpot CRM.
For each product it returns the reviews, newest first by default, and one summary row. You can keep only some star ratings, company sizes or lengths of use, or only the last N days. The site does the filtering, so you do not pay for reviews that are left out.
Capterra, Software Advice and GetApp: one review database
The three sites belong to the same company and show the same reviews. A review keeps the same ID (Capterra___ followed by a number) on Capterra and Software Advice, and a product has the same UUID on both.
This Actor reads the reviews from Software Advice, not capterra.com: capterra.com sits behind a Cloudflare challenge that plain HTTP requests do not pass, while Software Advice answers them. So productUrl in the results points to the product's reviews page on Software Advice, and productSeoId is the product's ID there, which is not its Capterra ID (Slack is 135003 on Capterra and 154669 on Software Advice).
How to scrape all Capterra reviews, not just the first 2,500
Capterra shows 25 reviews per page and stops at page 100. Ask for page 101 and it sends page 1 again, with a normal 200 response and no error. Software Advice goes to page 400, which is 10,000 reviews, then answers with an error. A scraper that pages through a big product gets a fraction of it, or pays for page 1 twice.
This Actor reads a product newest first until it hits the 10,000-review limit. If the product has more, it splits the rest into slices that the site's own filters cut:
- by overall rating, 1 to 5 stars, an exact cut;
- if a slice is still too big, by the reviewer's company size, A to I;
- what is left of a slice is read again by each length of use, and in both sort orders, until the slice's count is reached.
Since 5 November 2026 the site no longer filters on "not specified", so reviews without a company size can only be reached through those extra readings. Most are, not all. Measured on 6 November 2026, with Max reviews per product at 0:
| Product | Reviews on the site | Paging Capterra | Paging Software Advice | This Actor |
|---|---|---|---|---|
| Slack | 24,225 | about 2,500 | 10,000 | 23,893 (98.6%) |
On 18 October 2026, before that change, the Actor returned every review of Slack (24,229), Trello (23,636) and Asana (13,764). When a run falls short, the log says how many reviews it could not reach, and the product row gives the site's total in totalReviews. The full Slack run took 418 seconds. Products under 10,000 reviews need no split and come out in full.
Why scrape Capterra reviews?
- Competitive research. Keep the 1 and 2 star reviews of a competitor, from companies of the size you sell to, and read the
cons. Group byreviewerIndustryto see who struggles, and readreasonsForSwitchingto see why people left their old tool for it. - Product feedback. Export the reviews of your own product into a spreadsheet, a BI tool or a language model, with pros, cons and the overall comment in separate columns.
- Review monitoring. Run it on a schedule in New reviews only mode and be told when a competitor, or you, got a new review.
- Sales and marketing. Find which industries and company sizes use a product, pull quotes for a case study (5 stars,
timeUsedofMoreThan2Years), and see which reviews were incentivized. - Category comparison. One product row per product, with the average sub-ratings, the recommendation rate and the rating breakdown.
How to scrape Capterra reviews
From the Apify console
- Click Try for free. In Products, paste a Capterra, Software Advice or GetApp URL, or type a name such as
HubSpot CRM, one product per line. - Set Max reviews per product. The default is
100, and0collects every review. - Optionally keep only some star ratings, company sizes or lengths of use, or only the last N days.
- Click Start, then download the results as JSON, CSV or Excel.
From the API
curl -X POST "https://api.apify.com/v2/acts/alexmorain~capterra-reviews-scraper/run-sync-get-dataset-items?token=<YOUR_TOKEN>" \-H "Content-Type: application/json" \-d '{"products": ["https://www.capterra.com/p/135003/Slack/"], "maxReviewsPerProduct": 100}'
More examples: "Using the Capterra reviews API from your code".
Which fields should I fill?
| You want | Fill | Notes |
|---|---|---|
| The reviews of one or several products | products | URLs from Capterra, Software Advice or GetApp, or product names. |
| Every review of a product | maxReviewsPerProduct = 0 | The default is 100. |
| Only complaints | stars = [1, 2] | Filtered by the site, so you do not pay for the others. |
| Reviewers from companies of some size | companySizes | Codes A to I and Unspecified, listed below. |
| Only reviewers by length of use | timeUsed | FreeTrial, LessThan6Months, LessThan12Months, LessThan2Years, MoreThan2Years, Unspecified. |
| Only recent reviews | lastDays or publishedAfter | Whole days, 1 to 3650, or YYYY-MM-DD read as midnight UTC. |
| The site's own ranking instead of newest first | sortBy = recommended | Most complete reviews first. |
| Reviews only, no summary row | includeProductSummary = false | Saves $0.003 per product and gives a clean CSV. |
| Only reviews with their full text | fullReviewsOnly = true | Leaves out the summary-only reviews, which are then neither returned nor charged. |
| A feed of new reviews | reviewMode = new-only | Add webhookUrl to be notified. |
| Name your own feed | registryKey | Keeps one registry while you edit the filters. |
Reviewer company size codes: A self-employed, B 2-10, C 11-50, D 51-200, E 201-500, F 501-1,000, G 1,001-5,000, H 5,001-10,000, I 10,000+ employees, Unspecified for reviewers who did not say.
Input examples
Two products, 500 reviews each, only the 1 and 2 star reviews from companies of 51 to 1,000 employees:
{"products": ["https://www.capterra.com/p/135003/Slack/", "Asana"],"maxReviewsPerProduct": 500,"stars": [1, 2],"companySizes": ["D", "E", "F"]}
Every review of one product, for a full export:
{"products": ["https://www.softwareadvice.com/collaboration/slack-profile/"],"maxReviewsPerProduct": 0}
The last 30 days of three products, reviews only, no summary row:
{"products": ["HubSpot CRM", "Pipedrive"],"lastDays": 30,"maxReviewsPerProduct": 0,"includeProductSummary": false}
A weekly feed of new reviews, with a webhook:
{"products": ["Slack", "Asana", "Trello"],"reviewMode": "new-only","maxReviewsPerProduct": 0,"webhookUrl": "https://hooks.zapier.com/hooks/catch/..."}
Details that need a word
- A name is not always the right product. A typed name is searched on Software Advice. An exact name match wins, unless a product with at least ten times more reviews has the same words in its name:
Zoomis also the name of a product with 63 reviews, next to Zoom Workplace with 14,690, and the Actor picks the bigger one. The site also renames products (monday.comis listed as "monday AI Work Platform"), and third-party add-ons ("... Integration") are left out of the match. If the match is wrong, paste the Software Advice URL. - A Capterra URL is matched by ID when it can be. The Actor asks Capterra's regional sites (UK, Canada, Australia) for the product's UUID and finds the same product on Software Advice. Those sites answer only some of the time from Apify servers, so after two tries per site the Actor falls back to the product name in the URL. A GetApp URL is matched by the name in its address. The log shows the product chosen.
- Max reviews per product applies to each product.
0means no limit. The default is100, the console form is prefilled with20. Several URLs of one product are collected once. - Filters (
stars,companySizes,timeUsed) combine: a review has to match all three. An empty filter, or one that lists every value, filters nothing. - Summary-only reviews. Since November 2026 the site shows part of its reviews as a short card: rating, date, the reviewer's job title, company size and industry, and a one-paragraph summary. The full text stays behind a login. These rows come with
fullReview=false, the text insummary, andpros,consand the sub-ratings set tonull. SetfullReviewsOnlytotrueto leave them out. - Unspecified. The site no longer filters on
Unspecifieditself, so for that value the Actor reads the wider slice and keeps the matching reviews. Reviews left out this way are not charged. - Dates.
lastDaysandpublishedAftermake the Actor read newest first and stop at the first older review. If you set both, the more recent date wins. ApublishedAfterin the future is refused. With a date filter,sortByis ignored, because newest first is what allows the early stop.
What data can the Capterra reviews scraper extract? Every output field
Two kinds of rows, told apart by type: review and product. A field the site does not publish is null.
Review rows
| Field | Type | Description |
|---|---|---|
type | string | review |
reviewId, reviewNumber | string, integer | Review ID, a UUID since November 2026 (Capterra___7000001 before), and the number of an older ID, null for a UUID |
reviewKind | string | primary or historical for a full review, secondary for a summary-only one |
fullReview | boolean | false for a summary-only review, whose full text stays behind the site's login |
summary | string | The one-paragraph summary of a summary-only review |
benefits | string | What the reviewer gained from the product, when given |
productName, productId, productUrl | string | Product name, product UUID, and the product's reviews page on Software Advice |
title | string | Review title |
rating | number | Overall rating, 1 to 5 |
easeOfUseRating, customerSupportRating, valueForMoneyRating, functionalityRating | number | The four sub-ratings, 1 to 5 |
pros, cons | string | What the reviewer liked and disliked |
overallComment | string | Overall experience, in the reviewer's words |
reasonsForChoosing, reasonsForSwitching | string | Why the reviewer chose the product, and why they switched to it from another one |
vendorReply, vendorReplyDate | string | The vendor's public reply and its date, when the site gives it |
publishedAt | string | Review date, ISO 8601 UTC |
reviewerName | string | First name and last initial, for example Alex D.. null for an anonymous review |
reviewerAnonymous | boolean | The reviewer asked to stay anonymous |
reviewerJobTitle | string | Reviewer's job title, when the site shows it (mostly on summary-only reviews) |
reviewerIndustry | string | Reviewer's industry |
reviewerCompanySize, reviewerCompanySizeCode | string | Company size in words (51-200 employees) and the site's filter code, A to I |
reviewerValidated | boolean | The site validated the reviewer (LinkedIn or proof of use) |
timeUsed, timeUsedCode | string | How long the reviewer had used the product (More than 2 years) and the filter code (MoreThan2Years) |
usageFrequency | string | How often the reviewer uses it, for example Daily |
usageDuration | string | How long the reviewer has used it, in the site's newer wording, for example 3-5 years |
reviewSource, reviewSourceCode | string | How the review was collected, in words and as the site publishes it |
incentivized | boolean | The reviewer received a gift or a raffle entry |
editedVersions | integer | Number of earlier versions, when the reviewer updated the review |
scrapedAt | string | When the row was collected, ISO 8601 UTC |
reviewSource is one of organic, invited by the site, nominal gift, invited by the site, raffle entry, invited by the vendor, invited by the vendor, nominal gift. incentivized is true for the three gift and raffle values. An unknown source code is passed through in reviewSourceCode with a null label.
Product rows
One per product, unless you turn it off with includeProductSummary.
| Field | Type | Description |
|---|---|---|
type | string | product |
productName, productId, productSeoId | string | Product name, UUID, and ID on Software Advice |
productUrl, logoUrl | string | The product's reviews page on Software Advice, and its logo |
categories | array | Software categories the product is listed in |
totalReviews | integer | Total reviews of the product on the site |
matchingReviews | integer | Reviews that match your filters (stars, company size, length of use) |
reviewsCollected | integer | Reviews this run delivered for the product |
overallRating | number | The product's average rating, 1 to 5 |
easeOfUseRating, customerSupportRating, valueForMoneyRating, functionalityRating | number | The product's average sub-ratings |
recommendPercent | number | Share of reviewers likely to recommend the product, in percent |
ratingBreakdown | object | Number of reviews per star rating, keys 5 to 1 |
prosSummary | array | The site's summary of what reviewers like |
consSummary | array | The site's summary of what reviewers dislike |
input | string | The entry of your input that led to this row |
scrapedAt | string | When the row was collected, ISO 8601 UTC |
To see whether a run read everything, compare reviewsCollected with matchingReviews. In all mode, with no review limit and no date filter, they are equal when the whole product was read. A page that still fails after several retries is logged, with how many reviews may be missing.
Example output
Illustrative rows, made-up values:
[{"type": "review","reviewId": "Capterra___7000001","productName": "Example Chat","productId": "00000000-0000-4000-8000-000000000000","title": "Easy to use across devices","rating": 5,"easeOfUseRating": 5,"customerSupportRating": 4,"valueForMoneyRating": 5,"functionalityRating": 5,"pros": "Switching between laptop and phone just works.","cons": "Replies in threads do not always notify you.","vendorReply": null,"publishedAt": "2026-10-04T17:40:11Z","reviewerName": "Chris S.","reviewerAnonymous": false,"reviewerIndustry": "Construction","reviewerCompanySize": "51-200 employees","timeUsed": "More than 2 years","usageFrequency": "Daily","reviewSource": "organic","reviewSourceCode": "NoIncentive","incentivized": false,"scrapedAt": "2026-11-03T09:00:00Z"},{"type": "product","productName": "Example Chat","productSeoId": "100001","categories": ["Team Communication", "Collaboration"],"totalReviews": 1200,"matchingReviews": 1200,"reviewsCollected": 1200,"overallRating": 4.6,"ratingBreakdown": {"5": 840, "4": 280, "3": 50, "2": 20, "1": 10},"recommendPercent": 86.1,"input": "Example Chat"}]
The Output tab has two ready views, Reviews and Products.
How to export Capterra reviews to CSV or Excel
From the run's Storage tab, download the dataset as CSV, Excel, JSON, XML or HTML. Through the API, add the format to the dataset URL: https://api.apify.com/v2/datasets/<defaultDatasetId>/items?format=csv, or format=xlsx for Excel.
For a clean spreadsheet, set includeProductSummary to false: otherwise product rows come in the same file with their own columns. Or add view=overview to the URL for twelve review columns: product, date, rating, title, pros, cons, industry, company size, length of use, incentivized, vendor reply and review ID.
Monitor Capterra reviews: New reviews only mode
Set Reviews to return to new-only and schedule the Actor. For each product it keeps a registry of every review it delivered, in a named key-value store of your account called capterra-reviews-registry. A review already in the registry is skipped and not charged, and a review enters the registry only after it was pushed to your dataset.
- First run. Set Max reviews per product to
0and leave the date filters empty. The Actor reads the whole history and, once a product was read to the end, stores the date as its "complete read" marker. - Later runs. The Actor reads only the reviews newer than the marker minus 60 days. The margin covers reviews the site publishes late, after moderation, with an older date. Reviews in the margin are read again and skipped.
- A run cut short (review limit, cost limit, timeout, unreadable pages) clears the marker, so the next run reads the full history again and fills the gap. Delivered reviews stay uncharged.
The registry is tied to the product and to the filters that change which reviews qualify (stars, companySizes, timeUsed), so feeds with different filters never mix. Set a Delta registry key to name a feed yourself; if you then change its filters, the Actor reads the full history once more, still without charging delivered reviews. The product summary row is charged on every run, so turn it off with includeProductSummary for a feed.
Webhook at the end of the run
Add a webhookUrl and the Actor POSTs a JSON summary when the run ends:
{"actor": "capterra-reviews-scraper","runId": "<run id>","datasetId": "<dataset id>","mode": "new-only","status": "succeeded","products": 3,"reviews": 12,"skippedAlreadySeen": 340,"targets": 3,"failedTargets": 0,"chargeLimitReached": false}
status is succeeded, partial (some products failed) or failed (all of them did). In Zapier, Make or n8n, filter on reviews above 0 to hear about new reviews only, then read the dataset with datasetId. The URL is stored as a secret and never logged. Redirects are not followed, except 307 and 308 (same POST, once). A webhook that does not answer is logged as a warning and never fails the run.
Using the Capterra reviews API from your code
Every run is an API call. Your token is in the Apify console under Settings, Integrations.
Python
from apify_client import ApifyClientclient = ApifyClient("<YOUR_TOKEN>")run = client.actor("alexmorain/capterra-reviews-scraper").call(run_input={"products": ["https://www.capterra.com/p/135003/Slack/"],"maxReviewsPerProduct": 500,"stars": [1, 2],})for row in client.dataset(run["defaultDatasetId"]).iterate_items():if row["type"] == "review":print(row["publishedAt"], row["rating"], row["title"], "|", row["cons"])
JavaScript
import { ApifyClient } from 'apify-client';const client = new ApifyClient({ token: '<YOUR_TOKEN>' });const run = await client.actor('alexmorain/capterra-reviews-scraper').call({products: ['Asana', 'Trello'],lastDays: 30,maxReviewsPerProduct: 0,});const { items } = await client.dataset(run.defaultDatasetId).listItems();for (const row of items.filter((r) => r.type === 'review')) {console.log(row.productName, row.rating, row.title);}
cURL
For a long run, such as a full export of a big product, start the run and poll it, since the synchronous endpoint times out after 300 seconds:
curl -X POST "https://api.apify.com/v2/acts/alexmorain~capterra-reviews-scraper/runs?token=<YOUR_TOKEN>&maxTotalChargeUsd=25" \-H "Content-Type: application/json" \-d '{"products": ["Slack"], "maxReviewsPerProduct": 0}'
maxTotalChargeUsd is a ceiling: the Actor stops cleanly when it is reached. We tested it on 19 October 2026: a run capped at $0.012 on the Bronze plan stopped after 7 reviews.
Capterra reviews scraper pricing
You pay per result, plus a small start fee, with no monthly fee. The price goes down with your Apify plan, which is why the Store shows a "from" price.
| Event | When it is charged | Free | Bronze | Silver | Gold and above |
|---|---|---|---|---|---|
| Review | Per review row pushed to the dataset | $0.001 | $0.0009 | $0.0008 | $0.0006 |
| Product | Per product summary row, which you can turn off | $0.003 | $0.0027 | $0.0024 | $0.002 |
| Actor start | Once when the run starts, per GB of memory (the default is 512 MB, so one event) | $0.005 | $0.005 | $0.005 | $0.005 |
Per 1,000 reviews: $1.00 on Free, $0.90 on Bronze, $0.80 on Silver, $0.60 from Gold.
Worked examples, with the summary row on:
- The prefilled run, 20 reviews of one product, on Free: 20 x $0.001 + $0.003 + $0.005 = $0.028. The default run of 100 reviews: $0.108.
- 1,000 reviews of one product: $1.008 on Free, $0.607 on Gold.
- The 1 and 2 star reviews of Slack, 175 according to the site's rating breakdown, on Free: $0.183.
- A full export of Slack, 24,229 reviews: $24.24 on Free, $21.81 on Bronze, $14.54 on Gold.
- A weekly watch of ten products with five new reviews each, on Free: 50 x $0.001 + 10 x $0.003 + $0.005 = $0.085. A week with nothing new costs $0.035, or $0.005 with the summary row off.
Reviews skipped in New reviews only mode, and reviews your filters leave out, are never charged. Set Max charge per run in the run options to cap the cost.
Why this Capterra reviews scraper
Plain HTTP, no browser, no login. The Actor calls the endpoint the Software Advice website uses to load more reviews. No browser means 512 MB is enough, and no account of yours is involved. There is no proxy by default: on 17 and 18 October 2026 the source answered Apify datacenter servers without one. The Proxy field is there if your runs get blocked.
It copes with the site's quirks. The endpoint's ID changes when the site is redeployed: the Actor reads the new one from the site's scripts when the stored one is refused. A review page that fails is retried, and a failing product does not stop the others.
Failure that tells you what to do. A run where every product failed ends as FAILED with the reason in the status message, not as an empty success. An input the Actor cannot use (a rating outside 1 to 5, a date in the future) ends the run with a message naming the field. If only some products failed, the run succeeds with one log line per failure and the webhook says partial.
Limits and what this scraper does not do
- No G2. G2 is a different company's reviews and is not covered. We tested it on 17 October 2026: G2 blocks plain requests, and through an unblocker it returned the first page of reviews for every page number we asked, so it was left out.
- No reviewer job title. Capterra.com shows it, the data Software Advice publishes does not. Names are cut to first name and last initial, and empty for an anonymous review.
- Matching can fail. A name is matched by a search, and a Capterra URL by ID only when Capterra's regional sites answer. When nothing matches, the run logs a clear error for that product. A product with few reviews on Software Advice returns few: when we tested, Mailchimp had a single one there.
- Order. The first 10,000 reviews of a product come in the order you asked for, the rest slice by slice. Sort by
publishedAtafter the export. - A slice over 10,000. If one combination of rating, company size and length of use holds more than 10,000 reviews, the Actor collects the first 10,000 of it and logs a warning. Slack, the biggest product we tested, did not hit this.
- Speed. Reviews come 25 per page, four pages at a time, and products are read one after the other. A full Slack export took about seven minutes. The default run timeout is one hour: raise it for many big products.
- The site can change. The Actor depends on how Software Advice serves reviews today. If it breaks, open an issue.
FAQ
Can I scrape Capterra reviews?
Yes. Paste a Capterra product URL, press Start and download the reviews as JSON, CSV or Excel, with ratings, pros, cons, reviewer industry, company size and date.
Is there a Capterra API?
This Actor works as one. Start a run through the Apify API with a product URL or name, then read the reviews from the dataset as JSON or CSV. It needs no Capterra account or key.
How do I scrape all the reviews of a Capterra product, not only the first 2,500?
Set Max reviews per product to 0. Capterra stops listing a product after page 100, about 2,500 reviews, and the Actor gets past that by reading the product in slices the site filters itself. For Slack that gave 23,893 of 24,225 reviews in about seven minutes; a product under 10,000 reviews comes out in full.
Can I scrape Software Advice or GetApp reviews?
Yes. Paste a Software Advice or GetApp product URL. The three sites show the same reviews, and a Software Advice URL points at a product most exactly.
How much does it cost to scrape 1,000 Capterra reviews?
$1.00 on the Free plan, $0.60 on Gold, plus $0.005 for the start and $0.003 for the product summary row.
Can I scrape only the negative reviews, or only one company size?
Yes. Set stars to [1, 2], companySizes to the codes you want, or both. You only pay for the reviews you get.
How do I get only the new reviews since my last run?
Set Reviews to return to New reviews only, schedule the Actor, and use Max reviews per product 0 for the first run. See "Monitor Capterra reviews".
Is it legal to scrape Capterra?
Our Actors are ethical and do not extract any private user data, such as email addresses, gender, or location. They only extract what the user has chosen to share publicly. We therefore believe that our Actors, when used for ethical purposes by Apify users, are safe. However, you should be aware that your results could contain personal data. Personal data is protected by the GDPR in the European Union and by other regulations around the world. You should not scrape personal data unless you have a legitimate reason to do so. If you're unsure whether your reason is legitimate, consult your lawyers.
Integrations, exports and agents
Runs can be scheduled from the Apify console. Combine a schedule with New reviews only and a webhook and you get a call only when new reviews appear. An Apify webhook on the ACTOR.RUN.SUCCEEDED event posts the run and dataset IDs to your endpoint, so a pipeline can pull the reviews without polling.
The dataset is readable over the Apify API in JSON, CSV, XML, Excel, RSS and HTML. Apify's integrations push it to Google Sheets, Slack, Airtable, Zapier, Make, n8n and the usual storage targets without code. Anything that speaks the Apify API, including MCP-capable agents through the Apify MCP server, can start a run and read the dataset.
Support
Something wrong or missing? Open an issue in the Issues tab with the link to the run. A reproducible input gets a fix faster than a description. To call the Actor from your code, see the API tab, with endpoints and snippets prefilled.
Guide and use cases
Use cases for outbound and automation, the output fields, pricing and common questions are on the Capterra Reviews Scraper page of my site. The Reviews family page puts it side by side with the other scrapers of the same family.