Booking.com Reviews Scraper

Scrape guest reviews of any hotel on Booking.com. Give it hotel URLs and get one row per review: score, title, positive and negative text, reviewer country, traveler type, room type, stay dates and the hotel's reply. No login required.

What data can you extract?

• Review id, score (1-10), title, positive text, negative text and the review language
• Review date (ISO 8601)
• Reviewer name, country and country code, and traveler type (couple, solo, family, group)
• Room type, number of nights, check-in and check-out date, stay status
• Hotel response and helpful votes
• Hotel id, name and URL

About 50% of Booking.com reviews are score-only: the guest left a rating but no text, so title, positive and negative text are empty for them (roughly half have a positive or negative text, a quarter have a title). Set onlyWithText to skip those.

Use with AI Agents (MCP)

Connect this actor to any MCP-compatible AI client, including Claude Desktop, Claude.ai, Cursor, VS Code, LangChain and LlamaIndex.

{
  "mcpServers": {
    "apify": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://mcp.apify.com?tools=santamaria-automations/booking-reviews-scraper",
        "--header",
        "Authorization: Bearer <APIFY_TOKEN>"
      ]
    }
  }
}

Example prompt: "Use booking-reviews-scraper to get the 200 newest reviews of https://www.booking.com/hotel/ch/motel-one-zurich.html and summarise the main complaints."

How to use

{
  "startUrls": ["https://www.booking.com/hotel/ch/motel-one-zurich.html"],
  "maxReviewsPerHotel": 200,
  "sortBy": "newest",
  "languages": ["en", "de"],
  "travelerTypes": ["COUPLES"]
}

An empty input {} also works: it scrapes 50 reviews of one example hotel.

Input parameters

Parameter	Type	Default	Description
startUrls	array	example hotel	Booking.com hotel URLs (/hotel/<country>/<name>.html, any language suffix or query)
hotelIds	array	none	Numeric hotel ids, for hotels whose URL cannot be resolved
maxReviewsPerHotel	integer	100	1-100,000 reviews per hotel (an empty input runs a 50-review demo on the example hotel)
sortBy	string	newest	newest, oldest, highest or lowest
languages	array	all	Language codes such as en, de, fr
travelerTypes	array	all	COUPLES, SOLO_TRAVELLERS, FAMILIES, GROUP_OF_FRIENDS, BUSINESS_TRAVELLERS
keyword	string	none	Only reviews containing this text
onlyWithText	boolean	false	Skip score-only reviews (no title and no text). Skipped reviews are neither saved nor charged; maxReviewsPerHotel then counts saved reviews. A hotel scan stops after 5 x maxReviewsPerHotel raw reviews or at the end of its list

What you get

One row per review:

Field	Type	Description
review_id	string	Booking.com review id
hotel_id	integer	Booking.com hotel id
hotel_url	string	Hotel page URL (null when only a hotel id was given)
hotel_name	string	Hotel name (null when only a hotel id was given)
reviewer_name	string	Reviewer first name or alias
reviewer_country	string	Reviewer country
reviewer_country_code	string	Reviewer country code
traveler_type	string	Couple, Solo traveler, Family or Group
room_type	string	Booked room type
nights	integer	Number of nights stayed
checkin_date	string	Check-in date
checkout_date	string	Check-out date
stay_status	string	Stay status
review_date	string	Review date, ISO 8601 UTC
score	number	Review score, 1 to 10
title	string	Review title (only some guests write one)
positive_text	string	What the guest liked
negative_text	string	What the guest disliked
language	string	Language code of the review text
hotel_response	string	Reply from the hotel
helpful_votes	integer	Helpful votes
review_url	string	Reviews page of the hotel on Booking.com
scraped_at	string	Scrape time, ISO 8601 UTC
source_platform	string	Always booking.com

Output example

{
  "review_id": "e1d2eedc5e49225f",
  "hotel_id": 1388383,
  "hotel_url": "https://www.booking.com/hotel/fr/hotelf1-paris-porte-de-montmartre.html",
  "hotel_name": "HotelF1 Paris Saint Ouen Marché Aux Puces",
  "reviewer_name": "Maria",
  "reviewer_country": "Spain",
  "reviewer_country_code": "es",
  "traveler_type": "Couple",
  "room_type": "Cabrio Room with Private Bathroom",
  "nights": 4,
  "checkin_date": "2026-10-11",
  "checkout_date": "2026-10-15",
  "stay_status": "stayed",
  "review_date": "2026-10-20T22:20:47Z",
  "score": 8,
  "title": null,
  "positive_text": "Hotel on the outskirts of Paris, small but cosy rooms, fair value for money",
  "negative_text": "The receptionist was on a video call and ignored us at check-in",
  "language": "es",
  "hotel_response": null,
  "helpful_votes": 0,
  "review_url": "https://www.booking.com/reviews/fr/hotel/hotelf1-paris-porte-de-montmartre.html",
  "scraped_at": "2026-10-21T12:00:00Z",
  "source_platform": "booking.com"
}

Fields Booking.com does not publish for a review are null.

Pricing

Pay per result: one result event for every review saved, plus a one-time start event per run. Prices: review-start $0.001 per run, review-result $0.0003 per review ($0.30 per 1,000 reviews).

Event	When
review-start	Once per run, right before the first review is saved
review-result	Per review saved to the dataset

Billing follows what is saved: only rows pushed to the dataset are charged, so score-only reviews dropped by onlyWithText cost nothing. Duplicates are never saved or charged twice, runs that find no reviews are not charged, and a restarted run does not charge reviews it already saved.

Performance

• Speed: 25 reviews per request, about 1 second per page, so 1,000 reviews take roughly a minute
• Traffic: about 0.8 MB per 1,000 reviews

Tips

• Hotel URLs are matched to Booking hotel ids by name. If a URL cannot be resolved the run says so; add the numeric id to hotelIds instead.
• The languages filter matches the language of the review text; reviews with a score only may have an unknown language code.
• sortBy: newest is the best choice for scheduled monitoring runs.

Related actors

• Booking.com Scraper - hotel listings with prices, ratings and availability for a destination.
• TripAdvisor Scraper - hotel and restaurant reviews from TripAdvisor.
• Google Maps Scraper - hotel listings and reviews from Google Maps.
• Trustpilot Scraper - company reviews from Trustpilot.

Data protection

This actor returns only data that is publicly visible on the source pages. Results can include personal data, such as names, phone numbers, email addresses and addresses of private persons, sole traders or contact persons. Where the source labels an entry type, it is included so you can filter the results yourself.

You are responsible for using the data lawfully, including the GDPR/DSGVO, national data-protection laws and unfair-competition rules (e.g. UWG) for marketing and outreach.