Airbnb Reviews Scraper

Collect every guest review from any Airbnb listing — rating, comment, date, reviewer name, host reply — in one table you can open in Excel, Google Sheets, or a BI tool.

No login. No API key. No Airbnb account.

Built for STR hosts, conciergeries, and revenue managers who need structured guest feedback without scrolling listing pages.

---

Also available

The Store does not cross-sell this suite. Pick the Actor that matches the job:

Actor	Use when
Airbnb Pro Host	Business emails and phones by city
Airbnb Property Details	Amenities, GPS, photos, completeness
Airbnb Calendar	Occupancy and day-by-day availability
Airbnb Price Monitor	Nightly rates across a market
This Actor — Reviews	Guest comments by listing URL
Airbnb Host Portfolio	Map a host’s listings from profile URLs

---

Who is this for?

You are…	Typical goal	Suggested setup
STR host / owner	Reputation watch on your listing	One listing URL
Conciergerie / PM	Quality checks across a portfolio	Several listing URLs
Revenue manager	Compare guest sentiment vs competitors	Two listings in one run
Analyst / journalist	Structured comments with dates	English localization via JSON

---

Quick start

1. Paste one Airbnb listing URL under Listing URLs.
2. Click Start.
3. Download CSV from the Dataset tab (Reviews view).

Store Try uses a real listing and 1 review page (~10 comments) so the daily health check finishes under 5 minutes. Omit maxReviewPages in API JSON to export the full listing (capped at 200 pages).

---

Ready-made examples (published tasks)

Example	Best for
Guest reviews for a Paris Airbnb listing	STR hosts monitoring Paris reputation
Guest reviews for a Lyon Airbnb listing	Conciergerie / PM quality checks in Lyon
Airbnb reviews for a Barcelona listing	International STR analysts
Bulk Airbnb reviews — two listings	Compare two properties’ reputation

---

What you get

Column	What it means for you
listingId	Airbnb’s numeric listing id — handy if you merge with other data
listingUrl	The exact link you pasted — easy to match rows back to your list
listingIndex / listingsTotal	Which listing in your batch (1 of 5, 2 of 5, …) when you run several URLs
reviewText	The guest’s comment (in the export language you choose — see below)
reviewTextOriginal	How the guest wrote it originally, when Airbnb exposes it — useful when the site translated the main text
rating	Star rating as a number
localizedDate	Month and year you can read at a glance (e.g. “March 2025”)
publishedAt	Precise timestamp when available
reviewerName	First name when shown
hostResponse	Your (or the host’s) public reply, if any
reviewId	Stable id when Airbnb provides it
scrapedAt	When this row was written

Need extra columns? In the JSON input, set "includeVerboseReviewFields": true to add things like reviewer region, highlight tags, and language flags — handy for deep analysis, heavier files.

Dataset views in Apify: use the Reviews (default export) view for a compact column set. With Include all extra fields (JSON), the dataset still stores every key — open Apify’s built-in All fields tab (no separate custom view needed).

---

How to run

1. Open Input.
2. Under Listing URLs, enter one link per line (or paste a JSON array of strings — not { "url": "…" } objects).
3. Click Run.
4. When it finishes, open Dataset and download CSV / JSON / Excel.

Example input:

{
  "listingUrls": [
    "https://www.airbnb.fr/rooms/25497233"
  ],
  "maxReviewPages": 1
}

You can also pass a single URL as "url": "https://www.airbnb.com/rooms/…" in JSON.

Parameter	Type	Default	Description
listingUrls	string[]	—	One /rooms/ URL per line
maxReviewPages	integer	all (max 200)	Try prefill: 1. Omit in API for the full listing
maxListings	integer	—	Cap how many URLs to process (JSON)
listingConcurrency	integer	3	Parallel listings in API mode (1–8)
localizedLanguage	string	site TLD	e.g. "en" for one language (JSON)

Output sample

{
  "listingId": "25497233",
  "listingUrl": "https://www.airbnb.fr/rooms/25497233",
  "rating": 5,
  "reviewText": "Great stay, quiet and central.",
  "localizedDate": "March 2025",
  "reviewerName": "Marie",
  "hostResponse": "Thank you!",
  "scrapedAt": "2026-08-20T08:00:00.000Z"
}

Download JSON, CSV, Excel, or HTML from the Dataset tab.

---

Language & dates

• Default: text and month names follow each listing’s site (e.g. airbnb.fr → French, airbnb.de → German). You don’t have to configure anything.
• Want everything in one language? In JSON, add e.g. "localizedLanguage": "en" so comments and date labels are aligned in English across all listings.
• Prefer the guest’s own wording in the main text column? In JSON: "preferOriginalReviewText": true.

Translations come from Airbnb’s own content — this actor does not call a separate translation API.

---

Power users (JSON input)

The form stays simple on purpose. Open the JSON editor when you need to:

• Cap or tune the run — maxListings, maxReviewPages, maxReviewsPerListing, reviewsPerPage
• Speed vs. caution — listingConcurrency (parallel listings; lower it if Airbnb throttles)
• Logs — logAllListingUrls to print every URL in the run log
• Fallback control — reviewsFeedDisabled / useListingPagesOnly to force page-based collection only
• Stays / retries — check-in/out, guests, timeouts, maxRequestRetries, etc.

See CHANGELOG.md for version history and behaviour changes.

---

While the actor runs

Progress appears in the run log. A copy is also saved under Storage → Key-value store → RUN_LOG, so you can review what happened without scrolling the whole console.

---

FAQ

• Official Airbnb API? No. The Actor uses the same public listing / reviews feed a visitor sees.
• Empty dataset? The listing ID must be a real /rooms/ URL. The old placeholder id does not return reviews.
• Only 10 reviews? Store Try sets maxReviewPages: 1. Omit it in JSON for the full history.
• Wrong language? Set "localizedLanguage": "en" (or another code) in JSON.

Local development

npm install
# Edit input.json at the project root
apify run
npm test

Full suite: npm run test:all — see tests/README.md.

Support

Something wrong? Check RUN_LOG and the run output.

The form uses a simple URL list (strings). If you still see old inputs shaped like { "url": "…", "0": "h", … }, the actor accepts those too (url first, or rebuild from numeric keys).

Custom integrations or consulting: corentin@outreacher.fr.