Tripadvisor Hotels Scraper

Search Tripadvisor for hotels in any destination and get a clean row per property: rating, review count, ranking, coordinates, full address, phone, amenities, price range, and live per-provider room offers.

Give it a destination name — Barcelona, Austin, Texas, Kyoto — or paste a Tripadvisor Hotels URL. No login, no browser, no API key.

What you get

One record per hotel:

Field	Example
name, url, locationId	Hotel Astoria, .../Hotel_Review-g187497-d228590-..., 228590
rating, reviewCount	4.3, 2039
rankingPosition, rankingOutOf, rankingString	80, 534, #80 of 534 hotels in Barcelona
hotelClass, propertyType, category	3.0, Hotel, HOTEL
latitude, longitude	41.393997, 2.153488
street, city, state, postalCode, country, addressString	Carrer De Paris, 203 … 08036 Barcelona Spain
neighborhoods	["L'Antiga Esquerra de l'Eixample", "Eixample"]
phone	+34 932 09 83 11
amenities	["Pool", "Restaurant", "Bar/Lounge", "Fitness center"]
priceRangeMin, priceRangeMax	158, 299
lowestPrice, lowestPriceValue, currency	$137, 136.87, USD
offers	per-provider rates with numeric price, currency, availability, free-cancellation date
aiReviewSummary	Tripadvisor's generated summary of what reviewers say
featuredReview	highlighted review snippet with author, rating, date, link
image, thumbnail	full-size and 300×200 photo URLs

Input

{
  "locationQueries": ["Barcelona"],
  "maxItems": 100,
  "includeSponsored": false,
  "proxyConfiguration": {
    "useApifyProxy": true,
    "apifyProxyGroups": ["RESIDENTIAL"],
    "apifyProxyCountry": "US"
  }
}

• locationQueries — destination names. Each is resolved to a Tripadvisor destination, and the match is reported back on every row as searchGeoId / searchGeoName so you can confirm what was searched. Tripadvisor's lookup is fuzzy and will happily match something for a nonsense query, so check searchGeoName if a result set looks wrong.
• startUrls — Tripadvisor Hotels list URLs, e.g. https://www.tripadvisor.com/Hotels-g187497-Barcelona_Catalonia-Hotels.html. Use these when you want a specific filtered or sorted list. Non-Hotels URLs are skipped with a warning.
• maxItems — cap per destination, default 100, hard ceiling 10000. 0 means zero hotels; it never means unlimited.
• includeSponsored — Tripadvisor injects the same ~7 paid placements at the top of every page of a destination. They are dropped by default so you are not charged for duplicated ads. Set true to keep them, flagged as isSponsored.

Notes and limits

• Residential proxies are required. Tripadvisor is behind DataDome. Apify datacenter proxies are blocked outright (HTTP 403 + captcha on every attempt); RESIDENTIAL works.
• Prices are for Tripadvisor's default one-night stay window. Every row carries checkInDate, checkOutDate, and pricesAreDefaultDates so the quoted rates are never ambiguous. Custom date ranges are not supported yet.
• Some fields are genuinely absent for unreviewed properties. Tripadvisor does not rank a property with zero reviews, publishes no curated amenity list for small ones, and sometimes has no photo. Measured on reviewed properties, rankingString is 100% populated and image 98.8%.
• A search that returns nothing still writes one row explaining why, rather than finishing with an empty dataset.

How it works

Every Tripadvisor /Hotels-g<geoId>-... page is server-rendered with an inline urqlSsrData blob holding the same GraphQL results the page's JavaScript would otherwise fetch. One HTML GET yields 30 hotels fully populated — no browser, no XHR round-trips. Pagination walks the -oa<offset>- slug, deduping on locationId because paid placements repeat on every page.

Requests use curl_cffi with Safari TLS impersonation; Chrome profiles are rejected at the JA3 stage. Sessions rotate automatically on 403.