Booking.com Hotel Search Scraper

Search Booking.com for a destination and date range and get back real hotel listings scraped from the live search results page: name, address, distance from center, price (when Booking.com shows one), review score, listing URL, and thumbnail image.

Why this uses a real browser

Booking.com puts an AWS WAF JavaScript challenge in front of non-browser clients. Plain HTTP requests (curl, httpx, even curl_cffi with Chrome TLS impersonation) all get an HTTP 202 challenge page with no real content -- confirmed directly during development. A real headless browser that executes JavaScript is required to pass the challenge and reach the actual results, so this actor runs headless Chromium via Playwright rather than raw HTTP.

Input

Field	Required	Description
query	yes	Destination, e.g. "Austin, TX"
checkin	yes	Check-in date, YYYY-MM-DD
checkout	yes	Check-out date, YYYY-MM-DD
adults	no	Number of adult guests (default 2)
rooms	no	Number of rooms (default 1)
maxResults	no	Max listings to return (default 25, max 100)
proxyConfiguration	no	Apify Proxy config (recommended -- Booking.com blocks most datacenter IPs even for browser traffic)

Output

One dataset item per hotel listing:

{
  "name": "The Driskill",
  "address": "604 Brazos Street, Downtown Austin, Austin, TX",
  "distanceFromCenter": "0.3 km from center",
  "price": 412.0,
  "priceAvailable": true,
  "priceRaw": "$412",
  "reviewScore": 8.9,
  "reviewScoreRaw": "8.9 Excellent",
  "url": "https://www.booking.com/hotel/us/the-driskill.html",
  "imageUrl": "https://cf.bstatic.com/...",
  "scrapedAt": "2026-08-13T18:02:11Z"
}

If Booking.com doesn't display a price for a listing on the search page (it happens), price is null and priceAvailable is false -- this actor never estimates a price it didn't actually see.

Known limitations

• Booking.com's anti-bot defenses are aggressive and can occasionally block even browser-driven requests; if that happens the run returns a single diagnostic item explaining what page was actually reached instead of silently returning nothing.
• Selectors track Booking.com's current data-testid markup as of this build; if Booking.com changes their frontend, extraction may need updating.