Airbnb Reviews Scraper

Pull every guest review for an Airbnb listing — full review text, star rating, author first name and avatar, trip date, host response — into clean structured JSON.

Whether you are a vacation-rental host benchmarking competitor properties, a sentiment-analysis team building a reputation dashboard, a content site publishing "what guests are saying," or an AI team assembling a hospitality RAG corpus, this actor gives you the reviews you need without managing scrapers, browsers, or CAPTCHAs.

---

What you can do with it

• Reputation analytics — monitor your listing's review velocity, rating distribution, and sentiment trends over time.
• Competitive benchmarking — see exactly what guests praise and complain about at neighbouring properties.
• Host-response audits — measure how often (and how quickly) hosts respond, and what they say.
• AI / RAG corpora — feed thousands of real travel reviews into a hospitality chatbot, recommendation model, or sentiment classifier.
• Travel content sites — surface "top quotes" pulled from real guests on your own pages.
• Fraud signals — flag listings with suspicious review patterns (velocity spikes, duplicate text, geographic clustering).

---

Why use this actor

• No Airbnb account required — runs on public data anonymous visitors can read.
• No CAPTCHAs, no browsers — point it at one or more listings and it returns the reviews.
• Up to thousands of reviews per listing — pagination is automatic.
• Original language preserved — both the original review text and Airbnb's localized translation when present.
• Stable JSON schema — fixed fields ready to feed straight into your warehouse, spreadsheet, or notebook.

---

How it works

1. You give the actor a list of Airbnb listing URLs (or numeric room IDs) and how many reviews to pull per listing.
2. For each listing the actor walks the public reviews endpoint Airbnb's own site uses, page by page, in your chosen sort order.
3. Each review is normalized to one row in the dataset, ready for download as JSON, CSV, Excel, or XML.

Pagination, retries, and proxy rotation are handled automatically — you only need to provide listings and a max-review cap.

---

Input

{
  "listing_urls": [
    "https://www.airbnb.com/rooms/983207483398233150"
  ],
  "max_reviews_per_listing": 100,
  "sort": "MOST_RECENT",
  "locale": "en",
  "proxy": {
    "useApifyProxy": true,
    "apifyProxyGroups": ["RESIDENTIAL"]
  }
}

Field	Type	Required	Description
listing_urls	string[]	one of urls/ids	Full Airbnb listing URLs (e.g. https://www.airbnb.com/rooms/12345678)
listing_ids	string[]	one of urls/ids	Numeric Airbnb room IDs
max_reviews_per_listing	integer	no	Hard cap per listing (default 100, max 5000)
sort	string	no	MOST_RECENT (default), BEST_QUALITY, RATING_DESC, RATING_ASC
locale	string	no	Two-letter language tag (default en) — Airbnb may auto-translate; the original language is preserved
max_concurrency	integer	no	Parallel listings (1–4, default 1)
proxy	object	no	Apify proxy configuration; residential recommended

---

Sample output

One record per review. Real shape of a single record:

{
  "id": "9876543210",
  "listing_id": "983207483398233150",
  "url": "https://www.airbnb.com/rooms/983207483398233150",
  "rating": 5,
  "language": "en",
  "comments": "Loved every minute of our stay. The hot tub at sunset is unforgettable. Anna was an excellent host — quick to respond, full of local recommendations.",
  "localized_review": null,
  "localized_response": null,
  "created_at": "2025-09-18T03:14:12Z",
  "localized_date": "September 2025",
  "localized_short_date": "Sep 2025",
  "trip_duration": "Stayed about a week",
  "highlighted_review_sentence": [
    "Anna was an excellent host"
  ],
  "reviewer_id": "147015926",
  "reviewer_name": "Marta",
  "reviewer_location": "Madrid, Spain",
  "reviewer_picture_url": "https://a0.muscache.com/im/pictures/user/Hosting-User-147015926/original/abcdef12-3456-7890.jpeg",
  "response": "Thank you Marta! We are so glad you enjoyed the sunsets — see you again soon.",
  "response_date": "2025-09-19T10:02:44Z"
}

---

Output schema

Field	Type	Description
id	string	Airbnb review ID
listing_id	string	Numeric room ID this review is for
url	string	Canonical listing URL
rating	integer	Guest star rating, 1–5
language	string	ISO language code of the original review (en, de, id, ...)
comments	string	Full review text in the original language
localized_review	string	Airbnb-localized version when the original was in another language
created_at	string	ISO 8601 timestamp
localized_date	string	Human-friendly "September 2025"
localized_short_date	string	Compact "Sep 2025"
trip_duration	string	"Stayed a few nights" / "Stayed about a week"
highlighted_review_sentence	string[]	Sentences Airbnb highlights in the UI
reviewer_id	string	Airbnb's numeric user ID
reviewer_name	string	Guest first name (Airbnb does not expose last names)
reviewer_location	string	Guest's hometown
reviewer_picture_url	string	Avatar URL
response	string	Host's reply text
response_date	string	ISO 8601 timestamp of the host's reply

---

Tips

• Use MOST_RECENT for fresh signal; switch to RATING_ASC to surface complaints first.
• Combine with the airbnb-scraper actor: scrape listings first, then feed the resulting id values into this actor for full reviews.
• For high-volume runs (many listings × thousands of reviews) split across multiple actor runs to keep each within memory limits and to spread proxy load.