Pull live nightly rates, ratings, and amenities from Google Hotels as JSON, call it as a REST API, or wire it into Claude over MCP.

I got tired of AI assistants recommending hotels with prices from their training data. A model can tell you the Ritz is nice. It cannot tell you what a room there costs this weekend. So I built a scraper that answers the second question.
The Google Hotels Scraper API is an Apify Actor that turns any Google Hotels search into structured JSON: one row per property, with live nightly and total prices, guest ratings, review counts, star class, amenities, images, and location data.
[[IM

G1]]
Live prices: nightly and total for your exact dates, in any ISO 4217 currency
Ratings: guest score, review count, and star class
Property details: amenities, location data, images, and a property token for follow-up lookups
Vacation rentals: flip one parameter to search short-term rentals instead of hotels
One honest limit up front: it returns prices and booking context, it does not book rooms.
Open the Google Hotels Scraper API page on Apify and sign in with a free account.
Set a query like "hotels in Lisbon", check-in and check-out dates, and guest counts. Add filters if you want: price range, star ratings, amenities.
Hit Run. Results land in a dataset you can read as JSON or download as CSV.
Pricing is pay per event: $0.02 setup per run plus $0.03 per results page. A one-page search costs about a nickel.
Every Apify Actor is callable over HTTP. One request runs the search and returns the rows:
curl "https://api.apify.com/v2/acts/johnvc~google-hotels-search-scraper/run-sync-get-dataset-items?token=YOUR_TOKEN" -X POST -H "Content-Type: application/json" -d '{"q": "hotels in Lisbon", "check_in_date": "2026-10-02", "check_out_date": "2026-10-04", "adults": 2}'This is the setup I actually use. Add the Actor as an MCP tool with this endpoint:
https://mcp.apify.com/?actors=johnvc/google-hotels-search-scraperThen ask Claude "find me a 4-star in Lisbon under $180 a night for the first wee

kend of October". It calls the scraper mid-conversation and quotes real properties with real prices. No scraping code, no glue.
Hotel revenue managers rate-shop their comp set on a nightly schedule. Analysts track what CES week does to Las Vegas room rates. Travel-app builders feed live prices into their products. AI builders give their travel agents actual numbers instead of vibes.
There are three one-click examples on the examples page: competitor rate shopping for your hotel, searching Google Hotels from Claude via MCP, and tracking Las Vegas hotel prices for CES week. The Peerlist project has the full field list.
No. Google does not offer a public hotels API. The scraper reads the public results and returns them as structured JSON, fresh at run time.
$0.02 setup per run plus $0.03 per results page, billed per event on Apify. A single-page search for one city and one date range lands around five cents, and a free Apify account is enough to try it.
Yes. Add the MCP endpoint above and the scraper shows up as a tool your agent can call. Ask for hotels in plain English; the agent fills in the parameters.
No. It returns prices, ratings, and booking context so you or your agent can decide. The booking itself happens on the hotel or OTA site.
I watch the Actor's issue tracker and usually respond within a day or two. If a field you need is missing, tell me there.
0
0
0