App Store Reviews API: All Countries, Filters, AI Summary

Get the latest Apple App Store reviews for any iOS app across many country stores in one run, from Apple's own public customer-reviews feed: star rating, title, review text, app version and date, plus the app's overall rating summary on every row. Filter by star rating, date and keyword, and optionally add an AI summary of complaints and feature requests.

• Who it's for: product managers and mobile teams after a release, ASO and growth teams, competitor research, support teams, sentiment analysis.
• Source: Apple's public review feed and iTunes Lookup/Search API. No login, no API key, no proxies; reviewer names are not returned.
• Why switch: bundle id input, "all" countries, no free-plan trial cap, and a status message that says what each input resolved to and why a run stopped.
• Output: flat JSON, CSV or Excel with the same fields every run; pay per review returned (see the Pricing tab).

Quick start (one click)

Open the Actor and press Start: the form is prefilled with this input.

{
  "apps": [
    "https://apps.apple.com/us/app/spotify-music-and-podcasts/id324684580"
  ],
  "countries": [
    "us"
  ],
  "maxResults": 100
}

Sample output (one row, more fields in the dataset):

appId	country	appName	developer	appAvgRating
570060128	us	Duolingo: Learn Languages	Duolingo	4.72262

Use cases

Post-release feedback check. Reviews since the release date, newest first.

{"apps": ["https://apps.apple.com/us/app/spotify-music-and-podcasts/id324684580"], "countries": ["us"], "since": "2026-10-01", "maxResults": 500}

Complaints only, for support and QA. 1- and 2-star reviews that mention a keyword.

{"apps": ["324684580"], "countries": ["us", "gb"], "ratings": ["1", "2"], "keywords": ["crash", "login"]}

Competitor research across markets. Reviews of a competitor in several countries at once.

{"apps": ["com.spotify.client"], "countries": ["us", "de", "fr", "jp"], "maxResults": 400}

Summary for a product meeting. Reviews plus an optional AI summary of themes (paid add-on, off by default).

{"apps": ["324684580"], "countries": ["us"], "maxResults": 200, "aiInsights": true}

How it compares

Checked against the most-used App Store review actors in the Apify Store (their Store pages, reviews and Issues tabs, 23 Oct 2026). Alternatives are described generically; every claim about them is visible on their public pages.

	This actor	Common in the alternatives
App input	URL, numeric id, bundle id or name; shows what each input resolved to	Id or URL; bundle id support is an open feature request on the most-used one; one alternative is reviewed as "always scrapes" its example app
Countries	Any storefront, or "all" (~175), skipping stores where the app is not sold	Usually one country per run or a fixed list
Apify free plan	No trial cap: you get every review Apple returns, pay per review	Two popular ones return only 10 reviews on the free plan (stated by their developers in reviews and issues)
When it stops early	Status message says why (maxResults, cost limit, Apple returned none) + per-country FEED-STATS	Users reported runs that "stop after 10 reviews" with no message
Output shape	Flat, the same fields every run, documented below	One user reported nested vs flat rows changing between runs
Filters	Stars, since date, keywords	Varies; often only date or none
Extras	Optional AI summary of complaints and requests	Rare
Depth	Up to ~500 reviews per app, country and sort order (Apple's public feed limit), both sort orders combined	Most have the same feed limit; at least one alternative uses a different Apple endpoint and reports more per country

Ready-to-copy input examples

All 1–2★ reviews mentioning a crash in five big markets:

{"apps": ["com.spotify.client"], "countries": ["us", "gb", "de", "fr", "jp"], "ratings": [1, 2], "keywords": ["crash", "bug"], "maxResults": 2000}

Every storefront where an app is sold, newest first, since a date:

{"apps": ["https://apps.apple.com/us/app/duolingo-language-lessons/id570060128"], "countries": ["all"], "since": "2026-09-01", "maxResults": 5000}

Compare competitors by name, with an AI summary:

{"apps": ["Duolingo", "Babbel", "Busuu"], "countries": ["us"], "maxResults": 1500, "aiInsights": true}

What does the App Store Reviews API do?

• Any app, any input: App Store URL, numeric app id (e.g. 284882215), bundle id (e.g. com.spotify.client) or the app name (exact title match preferred). The run status and FEED-STATS show what each input resolved to, so a wrong match is visible at once.
• Many countries per run, or all of them: us, gb, de, jp… or "all" for every App Store storefront (~175). Storefronts where the app is not sold are skipped and not charged; countries are fetched in parallel.
• Says why a run stopped: the status message tells you if it hit maxResults, your max cost per run, or Apple returned nothing for a country.
• Filters: only 1–2★ reviews (bug and churn signals), only reviews since a date, only reviews mentioning keywords like crash or login.
• App context on every row: app name, developer, current version, average rating and rating count.
• Sort: most recent (default) or most helpful.
• Robust against Apple's empty feed pages: Apple's feed often answers with empty pages that are not the end of the list. The actor reads all 10 pages, retries empty pages with other equivalent feed URLs (JSON and XML), adds the other sort order when one is thin, retries briefly on rate limits, and removes duplicates. Only if the feed gives nothing at all does it take the ~10 reviews shown on the public App Store page. Each row has a source field (rss:mostrecent, rss:mosthelpful or web-page), and the FEED-STATS record shows what was used per app and country.

Default input (the prefill)

{"apps": ["https://apps.apple.com/us/app/spotify-music-and-podcasts/id324684580"], "countries": ["us"], "maxResults": 100}

Output example (real row, text shortened)

{"appId": "570060128", "country": "us", "appName": "Duolingo: Learn Languages", "developer": "Duolingo",
 "appAvgRating": 4.72262, "appRatingCount": 5518508, "appVersion": "7.145.1", "reviewId": "14694367120",
 "date": "2026-10-21T11:55:11-07:00", "rating": 1, "title": "Constantly glitches",
 "text": "I have been using duo lingo now for over 300 days. In the first few months I had no issues…",
 "appVersionReviewed": "7.145.1", "voteSum": 0, "voteCount": 0}

Optional AI insights (aiInsights, off by default)

Set aiInsights: true to get one short analysis per run, saved as the AI-INSIGHTS record in the run's key-value store: sentiment split (positive / neutral / negative %), a 2–3 sentence summary, the top complaints and the top feature requests (each with an approximate share and a short example), plus exact rating counts computed from the data.

• Processed by Google Gemini. The review text is sent to Google's Gemini API (a Flash-Lite model). Google may use data sent through its free tier to improve its products and models, so do not turn this on for reviews you are not comfortable sharing with Google.
• Only text, rating and date are sent. Reviewer names, review IDs, titles and app metadata are not sent. Emails, phone-like numbers and @handles inside the text are masked first.
• One call per run, trimmed input: at most 200 of the newest reviews, each cut to 300 characters.
• Suggestions, not guarantees. AI output can be wrong or miss things; check it against the reviews in the dataset.
• If the AI call fails (rate limit, timeout, bad answer), you get the normal reviews and an error field in AI-INSIGHTS, and nothing extra is charged.
• Price: $0.02 per run, charged only when aiInsights is true, at least one review was found and the AI call succeeds. Off by default.

Honest limits

• About 500 reviews per app, country and sort order. Apple's public feed returns at most 10 pages × 50. With both sort orders a run can return up to ~1,000 unique reviews per app and country, but often fewer, because some feed pages stay empty. For more history, add countries or run it on a schedule and keep the dataset.
• If Apple returns nothing, the run still succeeds, the status message says which app/country was empty, and nothing is charged per review (only the start fee). Try again later or add countries.
• Only reviews with written text are in the feed (bare star ratings are not).
• appAvgRating and appRatingCount come from Apple's Lookup API for that country and can differ from what the App Store page shows.
• Minimal personal data: reviewer nicknames and profile links are deliberately not returned.

How much does it cost?

Pay per event: $0.10 per 1,000 reviews ($0.0001 per review) plus $0.005 per run start. 500 reviews cost about $0.055. Optional AI insights (aiInsights): $0.02 per run, only when the AI call succeeds. No subscription.

Data source and legality

Data comes only from Apple's public, login-free sources: the customer-reviews RSS feed (itunes.apple.com/{country}/rss/customerreviews/...), the iTunes Lookup/Search API and, only as a last resort when the feed is empty, the reviews embedded in the public App Store web page. No login, no API keys, no proxies. Review texts belong to their authors. Use them for analysis and follow Apple's terms and applicable privacy law if you republish them.

FAQ

Google Play reviews? Not included. This actor covers the Apple App Store only. Can I schedule it or call it by API? Yes: Apify schedules, the API, and integrations (Make, Zapier, Google Sheets, Slack). Why fewer than 500 reviews? Smaller apps or country stores have fewer written reviews, some feed pages stay empty on Apple's side, and filters reduce the output.

Use it from code, AI agents and automations

The same input as the form above works everywhere. Results are the run's dataset (JSON, CSV, Excel). You need an Apify API token (Console → Settings → API & Integrations).

Python (pip install apify-client)

from apify_client import ApifyClient

client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("stazis/appstore-reviews-api").call(run_input={
  "apps": [
    "https://apps.apple.com/us/app/spotify-music-and-podcasts/id324684580"
  ],
  "countries": [
    "us"
  ],
  "maxResults": 100
})
for item in client.dataset(run.default_dataset_id).iterate_items():  # apify-client < 3: run["defaultDatasetId"]
    print(item)

JavaScript / TypeScript (npm i apify-client)

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('stazis/appstore-reviews-api').call({
  "apps": [
    "https://apps.apple.com/us/app/spotify-music-and-podcasts/id324684580"
  ],
  "countries": [
    "us"
  ],
  "maxResults": 100
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();

HTTP, one call (waits for the run, up to 5 minutes, and returns the items)

curl -X POST "https://api.apify.com/v2/acts/stazis~appstore-reviews-api/run-sync-get-dataset-items" \
  -H "Authorization: Bearer $APIFY_TOKEN" -H "Content-Type: application/json" \
  -d '{"apps": ["https://apps.apple.com/us/app/spotify-music-and-podcasts/id324684580"], "countries": ["us"], "maxResults": 100}'

AI agents (MCP). This Actor runs with limited permissions, so AI agents can use it through the hosted Apify MCP server (Claude, Cursor, VS Code, Codex and other MCP clients). Add this to your MCP client config and sign in with Apify on first use:

{ "mcpServers": { "appstore-reviews-api": { "url": "https://mcp.apify.com?tools=stazis/appstore-reviews-api" } } }

Agents without an Apify account can pay per run in USDC through agi.apify.com (x402). Pricing is pay-per-event, so an agent pays only for the results it gets.

LangChain (pip install langchain-apify, set APIFY_TOKEN): from langchain_apify import ApifyActorsTool and tool = ApifyActorsTool("stazis/appstore-reviews-api") gives your agent this Actor as a tool.

No-code and scheduling: use the Apify app in Make, n8n or Zapier ("Run an Actor", then pick stazis/appstore-reviews-api), or create a Schedule in Apify Console to run it daily and send results to a webhook, Slack or Google Drive through Apify integrations.

Related actors by stazis

Greenhouse, Lever & Ashby Jobs API, SEC EDGAR Filings API, VIN Decoder API, Domain RDAP/WHOIS API.