Trip.com MCP Server avatar

Trip.com MCP Server

Pricing

from $10.00 / 1,000 mcp tool calls

Go to Apify Store
Trip.com MCP Server

Trip.com MCP Server

MCP server for Trip.com hotels: your AI agent can resolve a city, search hotels with nightly prices, guest scores and coordinates, read curated top-lists and fetch hotel detail. Works from Claude, Cursor or any MCP client; also runs as a one-shot hotel scraper.

Pricing

from $10.00 / 1,000 mcp tool calls

Rating

0.0

(0)

Developer

axly

axly

Maintained by Community

Actor stats

0

Bookmarked

1

Total users

0

Monthly active users

9 days ago

Last modified

Categories

Share

Connect an AI assistant to Trip.com hotel data. This Model Context Protocol (MCP) server exposes four tools that Claude, Cursor, ChatGPT-style agents, n8n or your own code can call: resolve a destination, search hotels for a city and dates, read Trip.com's curated hotel top-lists and fetch the detail of one hotel. The same Actor also runs as a normal one-shot hotel scraper that writes to a dataset.

No Trip.com login or API key is needed. Every tool call queries Trip.com when it is made, for the dates you ask for.

Tools

ToolWhat it returns
search_destinations(keyword)Matching Trip.com destinations: city_id, name, country, type, oversea flag, coordinates.
search_hotels(destination, check_in, check_out, adults, rooms, limit, oversea, currency)Hotels for a city (name or id) and dates: name, address, star class, guest score, nightly price with and without tax, coordinates, Trip.com page link. Up to 50 hotels per call, paged for you.
hotel_rankings(destination, oversea)Trip.com's curated top-lists for a city (for example "Luxury Hotels" or "Hot Spring Hotels") with their member hotel ids and list URL. Some cities have none.
get_hotel(hotel_id, city_id)One hotel: name, star class, city, country, opening year and highlight tags. It carries no prices; use search_hotels for those.

A tool call that fails or finds nothing returns an error field (or an empty list) and is not charged.

Connect it

Start the Actor in Standby mode (open it in the Apify Console and use its Standby tab, or call the URL directly). The server listens at

https://<your-username>--trip-com-mcp.apify.actor/mcp

Send your Apify API token as Authorization: Bearer <token>. In an MCP client that supports Streamable HTTP servers:

{
"mcpServers": {
"trip-com": {
"url": "https://<your-username>--trip-com-mcp.apify.actor/mcp",
"headers": { "Authorization": "Bearer <your Apify token>" }
}
}
}

Then ask things like "find 4-star or better hotels in Bangkok for 14-16 November and compare the guest scores".

Example tool output (search_hotels)

{
"count": 1,
"hotels": [
{
"hotel_id": 1617121,
"name": "InterContinental Tokyo Bay",
"address": "1 Chome-16-2 Kaigan, Minato Ward, Tokyo, 105-8576, Japan",
"city_id": 228,
"city_name": "Tokyo",
"star": 5.0,
"review_score": 4.6,
"score_scale": 5,
"price": 415.0,
"price_excl_tax": 332.0,
"currency": "USD",
"latitude": 35.653006,
"longitude": 139.762375,
"detail_url": "https://www.trip.com/hotels/detail/?hotelId=1617121&cityId=228",
"checkin": "2026-10-18",
"checkout": "2026-10-19"
}
],
"total_available": 4754,
"city_id": 228,
"city_name": "Tokyo"
}

price is the lowest nightly rate Trip.com displays for those dates, including taxes and fees; it can differ from the final booking price. review_score is out of score_scale (5).

Classic run (no AI needed)

Press Start with a destination and the Actor performs a one-shot hotel search and writes rows to the dataset with the same fields as above. Leave destination empty and the run ends immediately with a note (the MCP server only runs in Standby mode).

{
"destination": "Bangkok",
"checkIn": "2026-11-10",
"checkOut": "2026-11-13",
"adults": 2,
"maxItems": 50,
"currency": "EUR"
}
InputDefaultDescription
destinationemptyCity name or numeric Trip.com city id
checkIn, checkOut+7 / +8 daysYYYY-MM-DD, check-in not in the past
adults, rooms2, 1Guests (1-8) and rooms (1-4)
maxItems20Stop after this many hotels
currencyUSDISO 4217 code for prices

Schedule it or call it by API/webhook like any Actor to monitor rates over time.

Use cases

  • Give a travel or planning assistant live hotel prices and guest scores instead of stale training data.
  • Compare districts or cities by star class, score and price from inside a chat.
  • Build agentic travel workflows in n8n or Make with MCP tool nodes.

FAQ

Which clients work? Any MCP client that supports Streamable HTTP and custom headers (Claude Desktop through a connector, Cursor, MCP SDKs).

How fresh is the data? Each call reads Trip.com at call time for your dates.

Can it book rooms? No. It only reads public search and detail data.

What if a call returns an error? Trip.com occasionally blocks or changes a response; the tool returns an error field and you are not charged. Retry or narrow the request.

Is it legal? It returns publicly visible data. You are responsible for complying with Trip.com's terms and applicable law.