Google Hotels MCP Server
Provides hotel and vacation rental search capabilities for Google Hotels, returning rates, ratings, amenities, photos, coordinates, and detailed property information including price comparisons across booking sites.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Google Hotels MCP ServerFind hotels in Paris for March 15-20 under $200/night."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Google Hotels MCP Server
A hosted Model Context Protocol (MCP) server that gives Claude, Cursor, Windsurf and any other MCP client one Google Hotels tool. Search hotels and vacation rentals for a destination and a pair of dates, then open any property in full with the rate every booking site is charging, ratings, a review breakdown by topic, amenities, photos and coordinates, all as structured JSON, with no Google account and no Places API quota.
1,000 free credits every month, no card required, which is about 100 hotel searches.
https://mcp.hasdata.com/api/mcp?apis=google_travel_hotels
Contents
Related MCP server: AigoHotel MCP Server
What you need
An MCP client and a HasData API key from the dashboard, free to create with no card, and the free tier covers about 100 calls a month at the 10-credit rate. This is a remote server, so the simplest path is a URL and an x-api-key header, with no container to run and no Google Cloud project anywhere in the flow. A client that only speaks stdio reaches it through a thin launcher, published as @hasdata/google-hotels-mcp on npm and hasdata-google-hotels-mcp on PyPI, shown below.
Quick start
The server URL is the same for every client. We run it hands-on in Claude Code and Claude Desktop. The other blocks follow each client's own documented format for a remote server.
Field | Value |
URL |
|
Transport | HTTP, streamable |
Auth header |
|
Clients with OAuth support can add the same URL as a connector and sign in without putting a key in a config file.
claude mcp add --transport http google-hotels "https://mcp.hasdata.com/api/mcp?apis=google_travel_hotels" \
--header "x-api-key: HASDATA_API_KEY"Settings, then Connectors, then Add custom connector, then paste https://mcp.hasdata.com/api/mcp?apis=google_travel_hotels and sign in.
For the config-file route, Claude Desktop loads only local (stdio) servers, so it reaches a remote server through a stdio launcher. The @hasdata/google-hotels-mcp package is that launcher, and it reads the key from the environment. Add this to claude_desktop_config.json:
{
"mcpServers": {
"google-hotels": {
"command": "npx",
"args": ["-y", "@hasdata/google-hotels-mcp"],
"env": { "HASDATA_API_KEY": "YOUR_KEY" }
}
}
}For Python instead of Node, swap the launcher for the PyPI package, which uvx runs without a manual install:
{
"mcpServers": {
"google-hotels": {
"command": "uvx",
"args": ["hasdata-google-hotels-mcp"],
"env": { "HASDATA_API_KEY": "YOUR_KEY" }
}
}
}~/.cursor/mcp.json for every project, or .cursor/mcp.json for one:
{
"mcpServers": {
"google-hotels": {
"url": "https://mcp.hasdata.com/api/mcp?apis=google_travel_hotels",
"headers": { "x-api-key": "HASDATA_API_KEY" }
}
}
}~/.codeium/windsurf/mcp_config.json. Windsurf calls the field serverUrl, not url:
{
"mcpServers": {
"google-hotels": {
"serverUrl": "https://mcp.hasdata.com/api/mcp?apis=google_travel_hotels",
"headers": { "x-api-key": "HASDATA_API_KEY" }
}
}
}.vscode/mcp.json in the workspace:
{
"servers": {
"google-hotels": {
"type": "http",
"url": "https://mcp.hasdata.com/api/mcp?apis=google_travel_hotels",
"headers": { "x-api-key": "HASDATA_API_KEY" }
}
}
}Example prompts
Prompts, not code. Paste one in and the agent picks the tool itself. Each is annotated with the calls it takes, because every successful call costs 10 credits.
Find hotels in Barcelona for the nights of 12 to 15 November, two adults, and list the five cheapest with nightly rate, rating and class.
One call, 10 credits. Rates, ratings and amenities all come back together.
Same stay, four stars and up, free cancellation only, sorted by rating.
One call, 10 credits. Class, cancellation and sort order are filters on the one request.
Take the top result and show what each booking site charges for it, and whether that is a good price.
Two calls, 20 credits. The search hands back a propertyToken per property, and passing it in opens that property with every source Google compares, a typical price range and Google's own verdict on the deal.
Vacation rentals in Lisbon that week with at least two bedrooms and a pool.
One call, 10 credits, with vacationRentals on, bedrooms at 2 and the pool in amenity__.
Name the destination the way you would say it out loud. A bare Barcelona is read against wherever Google believes the search is running from, so a server in California answers with Californian hotels. Either write hotels in Barcelona or set gl to es, and the city is pinned.
Tools
Tool | What it returns |
| Properties with nightly and total rate, class, rating, review counts and a breakdown by topic, amenities, images, coordinates and nearby transit, alongside pagination and the brand tree. Given a |
One tool, read-only, covering both halves of the site. Without a propertyToken it searches. With one it opens a single property.
The samples below are trimmed from real calls, and hotel rates move daily. Read them as a shape.
A sample is the payload, not the whole response. A tools/call result carries one text block, and that text is itself JSON holding url, status, text and json, with the scraped data under json. From a raw JSON-RPC response the path is result.content[0].text, parsed, then .json. A chat client unwraps that for you and code talking to the endpoint directly does not.
Search hotels and vacation rentals
hasdata_google_travel_hotels_getGoogleHotels
Properties for a destination and a stay, with rates, ratings, amenities and images.
Parameter | Type | Required | Notes |
| string | yes | Destination, neighbourhood or hotel name. Phrase it as |
| string | yes |
|
| string | yes |
|
| number | Guest mix, 1 to 6 adults and up to 5 children, 6 guests in total | |
| string | Comma-separated ages, one per child, such as | |
| string |
| |
| number | Per night, in the selected currency | |
| string |
| |
| string | Star classes to keep, comma-separated, such as | |
| array | Property types such as | |
| array | Amenities such as | |
| string | Brand ids to keep. The search response carries the whole brand tree with its ids | |
| boolean | Narrow to properties carrying that flag | |
| boolean | Search rentals instead of hotels | |
| number | Minimums, rentals only | |
| string | Currency, and the country and language of the search | |
| string | Next page, taken from | |
| string | Switches the call from a search to one property in full |
Results arrive under properties, roughly twenty per page. Each property carries name, type, description, link, gpsCoordinates, ratePerNight and totalRate (each with the raw string and an extracted* number), hotelClass with extractedHotelClass, overallRating, reviews, locationRating, a ratings histogram, reviewsBreakdown with positive and negative mentions per topic, amenities, images and nearbyPlaces with walking and transit times. searchInformation.totalResults reports how many Google has, pagination carries nextPageToken, and brands lists the brand tree with the ids the brands filter takes.
Rates are quoted two ways.
lowestincludes taxes and fees,beforeTaxesFeesdoes not, and a property may carry only one of them. Compare like with like before ranking.
{
"name": "Casa Gràcia",
"propertyToken": "ChcI0r6-2uyF7rIuGgsvZy8xdmo2bnNxZhAB",
"type": "hotel",
"description": "Cozy quarters in a hip lodging with dining & a lively bar, plus a kitchen, a library & free Wi-Fi.",
"link": "https://room00hostel.com/barcelona/casa-gracia-hostel/",
"gpsCoordinates": { "latitude": 41.3974523, "longitude": 2.1593339 },
"ratePerNight": { "beforeTaxesFees": "US$61", "extractedBeforeTaxesFees": 61 },
"totalRate": { "beforeTaxesFees": "US$182", "extractedBeforeTaxesFees": 182 },
"hotelClass": "4-star hotel",
"extractedHotelClass": 4,
"overallRating": 3.9,
"reviews": 3811,
"locationRating": "4.6",
"amenities": ["Breakfast", "Air conditioning", "Airport shuttle", "Kid-friendly"],
"nearbyPlaces": [
{ "name": "La Pedrera - Casa Milà", "transportations": [{ "type": "Walking", "duration": "5 min" }] }
]
}One property in full
Pass propertyToken from any search result back into the same tool, keeping the dates, and the answer is that property alone. On top of the search fields it adds prices, the rate at every source Google compares, featuredPrices for the sponsored ones, typicalPriceRange for the stay, deal and dealDescription when Google marks the price as a good one, address, phone, directions, amenitiesDetailed grouped by category, excludedAmenities and otherReviews from Tripadvisor and the rest.
{
"prices": [
{
"source": "Booking.com",
"numGuests": 2,
"ratePerNight": { "lowest": "$61", "beforeTaxesFees": "$39", "extractedLowest": 61 }
}
],
"typicalPriceRange": { "lowest": "$51", "highest": "$68", "extractedLowest": 51, "extractedHighest": 68 },
"deal": "34% less than usual",
"dealDescription": "Great Deal",
"address": "Pg. de Gràcia, 116Bis, Gràcia, 08008 Barcelona, Spain",
"phone": "+34 931 74 05 28"
}Errors and failure paths
Your client almost never sees an HTTP error code from a tool call. The MCP layer answers 200 and puts the failure inside the result, with isError set to true and the reason as text. The agent reads a message where you might expect a status line.
A wrong key surfaces as tool output, not as a failed connection. tools/list accepts any non-empty key and returns the tool, so the client completes its handshake and shows green. The first tool call then comes back with isError: true and the text HasData API error: 401 Unauthorized. Watch for that string, because nothing earlier in the flow reports the problem.
A missing key is the one real HTTP error. Authorization runs before any tool, and the connection itself fails with 401. CORS headers are present, and a browser client reads the status and not an opaque network failure.
A missing date is rejected before it becomes a search. Drop checkOutDate and the call fails validation with 422, naming the field.
The dangerous failure is silent and geographic. A destination Google cannot place, and a bare city name read from the wrong country, both return 200 with a full page of properties somewhere else entirely. We searched Qwertyville and got eighteen real hotels near the datacentre the request left from. Check gpsCoordinates on the first result, or pass gl, before trusting a list.
Dates that make no sense are quietly repaired rather than refused. A check-out before the check-in, or a stay in 2020, still answers 200 with properties, because Google normalises the range instead of erroring. Validate the dates on your side if they come from a model.
Results that carry data also carry a requestMetadata.id worth quoting in support.
Pricing, free tier and limits
Every Google Hotels call costs 10 credits per successful call. Response size does not change the price, and opening one property costs the same as a search.
The free tier is 1,000 credits every month with no card, which is about 100 searches. It renews with the billing cycle, so a low-volume agent runs on the free tier indefinitely.
Paid plans start at $59 a month for 200,000 credits, which is 20,000 searches, or $2.95 per 1,000 searches. The unit price falls with volume to $0.83 per 1,000 on the largest plan, and annual billing takes ten months of the monthly rate for twelve. Current figures live on the pricing page.
Your plan also sets concurrency: 1 request at a time on the free tier, 5 on Startup, 15 on Basic, and 50 to 500 across the Growth tiers. Handle the overflow case defensively in anything unattended.
A request that comes back non-200 is not billed. Opening a property after a search is a second call, so budget for it.
Tool selection
The apis query parameter decides which tools your agent sees. Fewer tools means less context spent on tool definitions, and fewer chances for the model to reach for the wrong one.
?apis=google_travel_hotels the one tool in this repo
?apis=google_travel add Google Flights
?apis=google_travel_hotels,airbnb hotels plus Airbnb stays
?apis=google_travel_hotels,booking hotels plus Booking.comThe parameter takes provider names like google_travel and individual API names like google_travel_hotels. Misspelled names are ignored. If every name is wrong the request fails with 400, and the body lists both what it did not recognise and every valid value. Drop the parameter and the same endpoint exposes every HasData tool.
How it compares
Google has never opened a public Hotels API. Hotel Center is for property owners feeding rates in, and the Places API returns a business record with photos and reviews but no rate for a date range. The alternatives are affiliate APIs from individual booking sites, each covering its own inventory and each gated behind approval.
Google Places API | This server | |
Rates for a stay | Not offered | Nightly and total, per property |
What other sites charge | Not offered |
|
Reviews | Five review snippets | Counts, a star histogram and a breakdown by topic |
Vacation rentals | Not covered | The same tool with one flag |
Setup | Google Cloud project, billing, quota | One key and one URL |
Cost | Per request, after the free cap | Paid past the free tier, 10 credits a call |
What this server does not do. No booking and no payment. It reads rates, availability for the dates you ask about and the links Google itself points at, and hands the booking step back to you.
FAQ
Is there an official Google Hotels API?
No. Google runs Hotel Center for hoteliers publishing their own rates, and the Places API for business records, neither of which returns what a stay costs. This server reads the public results and returns them as structured JSON.
What is a Google Hotels MCP server?
A server that exposes Google Hotels as a tool an AI client can call. The client sends a tool call over the Model Context Protocol, the server fetches the properties and returns structured JSON, and the model works with the result. This one exposes a single tool and runs remotely.
Why did my search come back with hotels in another country?
Because a bare place name is resolved against Google's own idea of where the search is running, which is the datacentre the request leaves from. Write hotels in Barcelona instead of Barcelona, or set gl to the country code, and the destination sticks.
How do I see what each booking site charges?
Search first, then call the tool again with the propertyToken of the property you care about. The answer carries prices with one entry per source, typicalPriceRange for the stay and Google's deal verdict when the rate is unusually low.
Can I search vacation rentals?
Yes. Turn on vacationRentals and the same tool searches rentals, where bedrooms and bathrooms become useful filters.
Can I use this together with other HasData APIs?
Yes. The apis parameter takes a list, and ?apis=google_travel adds Google Flights alongside hotels. Drop the parameter and you get everything.
Compliance and personal data
HasData accesses publicly available data only. A platform's terms may restrict automated access, and you are responsible for your own compliance.
HasData links
Product page and request builder | |
Server documentation | |
Every HasData tool in one server | |
Client walkthroughs | |
Everything else we scrape | |
Plans and credit costs | |
Keys and usage | |
Node launcher on npm | |
Python launcher on PyPI |
Development
This repository is configuration and documentation for a remote server. There is no build step and nothing to containerize.
The tests in test/ assert the tool contract, the part that can break without a commit here. They check that ?apis=google_travel_hotels returns exactly one tool, that it still declares its required parameters, that the name has not changed, and that the key in use is actually accepted. That last check calls the tool for real and costs 10 credits, which is the price of a canary that can fail for the right reason.
# macOS and Linux
HASDATA_API_KEY=your_key_here npm test
# Windows PowerShell
$env:HASDATA_API_KEY="your_key_here"; npm testThe same suite runs in CI on every push and once a week on a schedule, because the upstream tool list can change without anyone touching this repository. A failure means the tool list moved, the key stopped working, or the endpoint was unreachable, and the assertion message says which.
Contributing
Corrections to the parameter table and the response sample are the most useful contribution, because those are the parts that drift. Include the call you made and the response you got. Pull requests from forks run the suite without a key, and the live checks skip instead of going red.
License
MIT. See LICENSE.
Available Tools
1 toolhasdata_google_travel_hotels_getGoogleHotelsgoogle_travel_hotels: GET /AInspect
Get Google Hotels Results
Searches Google Hotels for hotels and vacation rentals with check-in/check-out dates, guest mix (adults, children with ages), localization (domain, country, language, currency), and advanced filters: price range, property types, amenities, minimum rating, brands, hotel class, free cancellation, special offers, eco-certified, vacation rentals only, bedrooms, bathrooms, and sort order (lowest price, highest rating, most reviewed). Returns per-property name, description, link, GPS coordinates, check-in/out times, rate per night and total rate, price comparisons across sources, hotel class, overall rating, reviews, location rating, images, amenities, nearby places, deals, eco-certification, and pagination tokens. Use for travel research agents, hotel rate monitoring, OTA dashboards, and competitive analysis of accommodations across markets.
| Name | Required | Description | Default |
|---|---|---|---|
| q | Yes | The search query for Google Hotels (e.g., a city, neighborhood, or hotel name). | |
| gl | No | The two-letter country code for the country you want to limit the search to. Provide one exact documented value (245 allowed), e.g. `ac`, `af`. | |
| hl | No | The two-letter language code for the language you want to use for the search. Provide one exact documented value (159 allowed), e.g. `af`, `ak`. | |
| adults | No | Number of adult guests (1-6). | |
| brands | No | Comma-separated list of brand IDs to filter by. Brand IDs are returned in the response under `brands` for the same query. | |
| rating | No | Filter by minimum overall guest rating. Options: - `threePointFivePlus` — 3.5 stars or higher - `fourPlus` — 4.0 stars or higher - `fourPointFivePlus` — 4.5 stars or higher | |
| sortBy | No | Sort hotel results. Options: - `lowestPrice` — lowest price - `highestRating` — highest rating - `mostReviewed` — most reviewed | |
| bedrooms | No | Minimum number of bedrooms (vacation rentals only). | |
| children | No | Number of child guests (1-5). Total guests (adults + children) cannot exceed 6. | |
| currency | No | Parameter defines the currency of the returned prices Provide one exact documented value (71 allowed), e.g. `ALL`, `DZD`. | |
| maxPrice | No | Maximum price per night, in the selected currency. | |
| minPrice | No | Minimum price per night, in the selected currency. | |
| amenity__ | No | List of amenities to filter by (e.g., `hotelFreeWifi,hotelPool`). Values prefixed `hotel*` apply when searching hotels (the default mode). Values prefixed `rental*` apply when `vacationRentals=true`. The two sets are disjoint because Google Hotels exposes different amenity catalogs for each mode — pass values matching the mode you're querying. | |
| bathrooms | No | Minimum number of bathrooms (vacation rentals only). | |
| hotelClass | No | Comma-separated list of hotel star classes to include (e.g., `2,3,4,5`). | |
| checkInDate | Yes | The check-in date in 'yyyy-MM-dd' format. | |
| checkOutDate | Yes | The check-out date in 'yyyy-MM-dd' format. | |
| childrenAges | No | Comma-separated list of children ages (e.g., `5,8,12`). Must match the number of children. | |
| ecoCertified | No | Show only eco-certified properties. | |
| nextPageToken | No | Token to fetch the next page of hotel results. Returned in the `pagination` field of a previous response. | |
| propertyToken | No | Token used to retrieve details for a specific property. Returned in each property in the results. | |
| specialOffers | No | Show only properties with special offers. | |
| propertyType__ | No | List of property types to filter by (e.g., `hotelResort,hotelMotel`). Values prefixed `hotel*` apply when searching hotels (the default mode). Values prefixed `rental*` apply when `vacationRentals=true`. The two sets are disjoint because Google Hotels exposes different property-type catalogs for each mode — pass values matching the mode you're querying. | |
| vacationRentals | No | Search vacation rentals instead of hotels. | |
| freeCancellation | No | Show only properties offering free cancellation. |
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations were provided (no read-only or destructive hints), so the description carries the full burden. The description thoroughly explains what the tool does and returns, and as a search tool it is implicitly non-destructive. It doesn't mention rate limits or auth, but nothing contradicts the (absent) annotations, so no contradiction flag.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single dense paragraph followed by use cases — reasonably structured and free of filler. It's long, but the density of useful information (explicit use cases, return scope) justifies the length without becoming wordy.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
There is no output schema, so the description doesn't need to detail return values. It compensates by explaining the data model through parameter docs (brands, pagination, property tokens) and by stating the range of returned per-property fields. Adequate for an agent to invoke the tool correctly.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, and the descriptions add substantial value beyond the raw types: enum options (rating, sortBy) are human-readable, amenities and propertyType explain the disjoint hotel vs rental catalogs, childrenAges explains the count-matching requirement, and nextPageToken/propertyToken explain where the tokens originate. This far exceeds baseline expectation.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's purpose — searching Google Hotels for hotels and vacation rentals with dates, guest mix, localization, and filters. It distinguishes itself from sibling tools (e.g., Google Flights, Airbnb, Booking) by being specifically for Google Hotels data, and explicitly names its use cases.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explicitly lists when to use this tool — travel research agents, hotel rate monitoring, OTA dashboards, and competitive analysis. Combined with the sibling-tool list (Google Flights, Airbnb, Booking, etc.), it clearly routes the agent to the right tool without ambiguity.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
1 tool update
v1.0.0- First observed
hasdata_google_travel_hotels_getGoogleHotels
Related MCP Connectors
Hotel booking MCP server. Search, book, and manage reservations across 250K+ properties worldwide.
Search MCP servers, agent skills and APIs for travel: hotel ID mapping, rates, rooms.
Search hotel prices, get best overall and best direct price in structured response. Get your developer token at https://Infoseek.ai/mcp
Search hotels, get live prices, and check out in chat. Guest search needs no sign-in.
Related MCP Servers
- AlicenseAqualityDmaintenanceEnables AI agents to search, browse, and book hotels from a database of 2 million properties worldwide. Provides comprehensive hotel search capabilities with location lookup, filtering by amenities, detailed property information, and integrated booking functionality.66 npm1ISC
- AlicenseNot gradedqualityDmaintenanceEnables AI assistants to search for hotels globally via the MCP protocol, supporting filters like city, landmark, star rating, and distance.2MIT

DIDA Hotel MCPofficial
AlicenseNot gradedqualityBmaintenanceEnables AI agents to search, compare, and book hotels with real-time pricing and availability, supporting multiple location types, star ratings, and price filters.8MIT- AlicenseNot gradedqualityBmaintenanceMCP server for searching stays, comparing prices across Booking.com, Airbnb, Vrbo, and Google Hotels, and fetching reviews via natural language in AI assistants.MIT