Booking.com Scraper

Scrape Booking.com hotel and accommodation search results into clean JSON or CSV. Pick a destination and stay dates. For every property you get the price for your stay (total and per night, with currency), the original price and discount, review score and count, star rating, accommodation type, address, district, coordinates, distance from the center, cancellation and prepayment policy, room or unit details, a photo and a ready-to-book link.

Powered by ScrapeUnblocker: the getPageSource API loads the Booking.com results page in a real browser and returns the HTML. This package reads the data Booking.com embeds in that page and turns it into typed records. You don't need to run proxies, headless browsers or retry logic yourself.

Features

• Search like the website does. Any destination Booking.com's search box understands (Paris, New York, Lake Tahoe, Shibuya), with check-in/check-out dates, adults, children's ages, rooms, currency and site language. You can also paste a results URL from your browser and it is scraped with all its filters.
• Accurate prices. The total for the stay and the nightly rate in your chosen currency, plus the price before discount, the discount amount, deal badges such as Limited-time Deal, taxes and fees not included in the price, and the price in the property's own currency.
• Rich, typed records. Booking.com property ID, name, accommodation type, matched room or unit, bedrooms/bathrooms/beds and floor area for entire homes, star rating, review score, review word and count, location score, address, district, city, country, latitude/longitude, distance from the center, nearest public transport, free cancellation (and deadline), no prepayment, meal plan, "only X left" message, highlights, Preferred Partner, sustainability certification, sponsored flag, host type and photo URL.
• Every Booking.com filter. Shortcuts for star rating, minimum review score, property type, nightly price range, free cancellation, no prepayment and breakfast. The filters command lists every other filter Booking.com offers for a destination, such as amenities, neighborhoods, brands, landmarks, bed type and accessibility, with counts. You can pass any of them with --filter.
• Every sort order. Top picks, lowest/highest price, top reviewed, best reviewed and lowest price, distance from downtown, property rating, and more.
• More than one page of results. Booking.com shows 15-25 properties per request and ignores paging offsets on the server-rendered page. So the scraper sweeps: it re-runs your search with other sort orders (and, with --split-by-stars, one star class at a time) and de-duplicates by property ID until it reaches --limit. You can cap the cost with --max-requests.
• Resilient. Timeouts, transient API errors and pages without results data are retried with doubling backoff. An unknown destination raises a clear error instead of silently returning Booking.com's home page.
• JSON, JSON Lines or CSV. CSV joins list fields with "; " and is written as UTF-8 with a BOM so Excel shows accented hotel names correctly.
• CLI + Python library. Type-hinted, and tested offline with a mocked client.

Install

git clone https://github.com/ScrapeUnblocker/booking-scraper.git
cd booking-scraper
pip install .

Get an API key at scrapeunblocker.com and expose it as SCRAPEUNBLOCKER_KEY (the official SDK reads it from there):

export SCRAPEUNBLOCKER_KEY=your_key_here        # Windows PowerShell: $env:SCRAPEUNBLOCKER_KEY="your_key_here"

Or copy .env.example to .env and load it with your tool of choice. Never commit a real key.

CLI usage

# 25 properties in Paris for 2 nights, 2 adults (JSON to stdout)
booking-scraper search "Paris" --checkin 2026-11-10 --nights 2

# 4-5 star hotels rated 8+ under $200/night, cheapest first, to CSV
booking-scraper search "Lisbon" --stars 4,5 --min-review 8 --type hotels \
    --max-price 200 --sort price -o lisbon.csv

# A family of four, two rooms, prices in euros
booking-scraper search "Barcelona" --adults 2 --children-ages 5,9 --rooms 2 --currency EUR

# Collect 150 properties: sweep sort orders and star classes, at most 20 requests
booking-scraper search "Rome" --limit 150 --split-by-stars --max-requests 20 -o rome.csv

# See every filter Booking.com offers for a destination...
booking-scraper filters "Amsterdam" --checkin 2026-11-20
# ...and use any of them (Free Wifi + pet friendly)
booking-scraper search "Amsterdam" --filter hotelfacility=107 --filter stay_type=1

# Scrape a results URL copied from your browser
booking-scraper search --url "https://www.booking.com/searchresults.html?ss=Tokyo&checkin=2026-12-01&checkout=2026-12-04&group_adults=2&nflt=class%3D4"

python -m booking_scraper ... works too. Progress goes to stderr, so stdout stays clean for piping. Add -q to silence it.

Option	Meaning
destination / --url	Free-text destination, or a Booking.com searchresults URL
--checkin, --checkout, --nights	Stay dates. Check-in defaults to 30 days from today, 1 night
--adults, --children-ages, --rooms	Party size, e.g. --children-ages 5,9
--currency, --lang	Price currency (default USD) and site language (default en-us)
--dest-id, --dest-type	Pin an exact Booking.com destination id (as seen in search URLs)
--sort	popularity, price, price_from_high_to_low, bayesian_review_score, review_score_and_price, class, class_asc, class_and_price, distance_from_search, upsort_bh, business_popularity
--stars, --min-review, --type	Star classes, minimum score (6/7/8/9), property types (hotels, apartments, hostels, resorts, bed-and-breakfasts, villas, campgrounds, guesthouses, condo-hotels, vacation-homes, homestays)
--min-price, --max-price	Nightly budget in --currency
--free-cancellation, --no-prepayment, --breakfast	Policy and meal filters
--filter KEY=VALUE	Any filter from the filters command (repeatable)
--limit	Max properties (default 25). Sweeps sort orders when one request isn't enough
--no-sweep, --split-by-stars, --max-requests	Control how far a sweep goes
--no-sponsored	Drop sponsored placements (kept and flagged by default)
-o, --format	Output file (.json, .jsonl, .csv) or explicit format
--country, --retries	Request routing country (default US) and retries per request

Library usage

from booking_scraper import BookingScraper

scraper = BookingScraper()  # reads SCRAPEUNBLOCKER_KEY; or BookingScraper(api_key="...")

hotels = scraper.search(
    "Paris",
    checkin="2026-11-10",
    nights=2,
    adults=2,
    stars=[4, 5],
    min_review=8,
    free_cancellation=True,
    currency="EUR",
    limit=50,
)
for h in hotels:
    print(h.name, h.review_score, h.price, h.currency, h.booking_url)

Lower-level building blocks:

from booking_scraper import BookingScraper, SearchQuery, make_query

scraper = BookingScraper(retries=5, on_retry=lambda n, delay, exc: print("retry", n))

# One request -> one SearchPage with the search's own metadata
page = scraper.fetch_page(make_query("Lisbon", checkin="2026-11-20", sort="price"))
print(page.destination, page.dest_id, page.total_results, page.applied_filters)
print(page.sorts)  # sort orders available for this search
for group in page.filters:  # every filter, with property counts
    print(group.title, [(o.filter, o.label, o.count) for o in group.options][:3])

# Build a query from a browser URL, then tweak it
query = SearchQuery.from_url(
    "https://www.booking.com/searchresults.html?ss=Rome&checkin=2026-12-01"
)
cheap = scraper.search_query(query.with_options(sort="price", add_filters=["fc=2"]), limit=10)

# De-duplicate across several searches by sharing a seen-set
seen: set[int] = set()
coast = []
for town in ["Positano", "Amalfi", "Ravello"]:
    coast += scraper.search(town, checkin="2027-06-10", nights=3, limit=25, seen=seen)

Save results with booking_scraper.export.write(properties, "out.csv") (or .json / .jsonl).

Errors: DestinationNotFoundError when Booking.com can't match the destination, BookingParseError when a page held no results data after all retries, and the SDK's own exceptions (bad key, out of credit, ...) are passed through unchanged.

Example output

booking-scraper search "Barcelona" --checkin 2026-11-20 --nights 2 --limit 40 (real output, first record):

Searching Booking.com: Barcelona 2026-11-20 -> 2026-11-22
  Barcelona: 23 properties (sort=popularity, filters=-, 1258 total)
  Barcelona: 20 properties (sort=price, filters=-, 1257 total)
40 properties

{
  "property_id": 93521,
  "name": "Eco Boutique Hostal Grau",
  "url": "https://www.booking.com/hotel/es/hosteria-grau.html",
  "booking_url": "https://www.booking.com/hotel/es/hosteria-grau.html?checkin=2026-11-20&checkout=2026-11-22&group_adults=2&no_rooms=1&group_children=0&selected_currency=USD&lang=en-us",
  "position": 1,
  "accommodation_type": "Hotel",
  "accommodation_type_id": 204,
  "room_name": "Double or Twin Room",
  "unit_type": null,
  "bedrooms": null,
  "bathrooms": null,
  "beds": 2,
  "area": null,
  "area_unit": null,
  "star_rating": 2,
  "star_rating_type": null,
  "review_score": 9,
  "review_word": "Wonderful",
  "review_count": 1108,
  "location_score": 9.7,
  "external_review_score": 9.7,
  "external_review_count": 14,
  "address": "Ramelleres 27",
  "district": "Ciutat Vella, Barcelona",
  "city": "Barcelona",
  "country_code": "es",
  "latitude": 41.384690149815135,
  "longitude": 2.1681831032037735,
  "distance_from_center": "0.2 miles from downtown",
  "public_transport": "Plaça de Catalunya (FGC) station is within 150 yards",
  "price": 324.29,
  "price_per_night": 162.15,
  "currency": "USD",
  "price_before_discount": null,
  "discount": null,
  "taxes_and_fees": 32.42,
  "property_currency_price": 287.87,
  "property_currency": "EUR",
  "deals": [],
  "sold_out": false,
  "free_cancellation": false,
  "free_cancellation_until": null,
  "no_prepayment": false,
  "meal_plan": null,
  "availability_message": null,
  "highlights": ["Shuttle service", "24-hour front desk"],
  "preferred_partner": true,
  "sustainability_certified": true,
  "sponsored": false,
  "host_type": "professional",
  "photo_url": "https://cf.bstatic.com/xdata/images/hotel/square600/147973986.jpg?k=ccede9a6cde3f1f7cfcd73f47b788f8111b5b6a12b763bda018bb6fcd05ad4e3&o=",
  "checkin": "2026-11-20",
  "checkout": "2026-11-22",
  "destination": "Barcelona"
}

Prices and availability are live, so you'll see different numbers.

Field notes

• price is the total for the whole stay and party as Booking.com displays it. taxes_and_fees are the charges Booking.com lists as not included in that price (city tax, cleaning, ...).
• price_per_night is Booking.com's own average nightly rate as shown on the card. It isn't always price / nights, because Booking.com can leave some charges (such as one-off fees) out of it. When the page shows no nightly rate, it is price / nights.
• discount = price_before_discount - price, filled only when Booking.com shows a struck-through price.
• review_score is on Booking.com's 0-10 scale and is null when Booking.com hides it (too few reviews). external_review_score / external_review_count are the external-review figures Booking.com shows for some properties.
• bedrooms, bathrooms, area and unit_type are filled for entire homes and apartments. For hotels, room_name is the room Booking.com matched to your party.
• Distances and areas are the text Booking.com shows for the site language (en-us: miles and ft²).
• sponsored marks paid placements in the results. They are real, bookable properties matching your search, so they're kept by default. Use --no-sponsored to drop them.

Project layout

booking-scraper/
├── src/booking_scraper/
│   ├── __init__.py      # public API + __version__
│   ├── scraper.py       # BookingScraper: fetching, retries, sweeps, de-duplication
│   ├── parsing.py       # embedded results data -> Property / SearchPage / FilterGroup
│   ├── urls.py          # SearchQuery, filters, sort orders, URL building/parsing
│   ├── models.py        # dataclasses
│   ├── export.py        # JSON / JSON Lines / CSV writers
│   ├── cli.py           # `booking-scraper` command
│   └── __main__.py      # `python -m booking_scraper`
├── examples/
│   ├── search_to_json.py     # one search -> JSON file
│   ├── best_value_hotels.py  # 4-5 star, 8+ score, under a budget, cheapest first
│   └── city_to_csv.py        # sweep a whole city into CSV
├── tests/                    # offline tests with a fake client + saved page fixture
├── pyproject.toml
├── requirements.txt
├── Makefile
└── .github/workflows/ci.yml

Development

python -m venv .venv && source .venv/bin/activate
make install      # pip install -e ".[dev]"
make lint         # ruff check
make format       # ruff format
make test         # pytest (offline, no API key needed)
pre-commit install

The tests replay a saved, trimmed Booking.com results page and synthetic pages through a fake client. They spend no API credit and run in CI on Python 3.9 and 3.12.

Responsible use

Scrape at a reasonable pace and respect Booking.com's terms of use. Prices and availability change constantly, so treat scraped prices as a snapshot and confirm on Booking.com before you rely on them. This project is not affiliated with or endorsed by Booking.com.

Links

• ScrapeUnblocker - the web scraping API behind this project
• ScrapeUnblocker documentation - getPageSource, SDKs, parsing options
• Python SDK on PyPI - pip install scrapeunblocker

License

MIT - Copyright (c) 2026 ScrapeUnblocker