Expedia MCP Server

The Expedia Travel Recommendations MCP Server is an official open-source project from Expedia Group that exposes Expedia's travel recommendation APIs through the Model Context Protocol. It allows LLMs and AI agents to query Expedia for hotel, flight, activity, and car rental recommendations using a standardized interface, supporting both stdio (for local LLM integration) and streamable-http (for web-based applications) transports.

The server wraps four primary recommendation endpoints (/expedia/hotels, /expedia/flights, /expedia/activities, /expedia/cars) and accepts structured search parameters like destination, dates, property types, amenities, guest ratings, and sort preferences. It is built in Python 3.11+ and distributed as a Python package runnable via uvx, as well as a Docker image (expediagroup/expedia-travel-recommendations-mcp).

Note that Expedia's broader AI/MCP program is described on their developer hub as still in an "exploration phase." This MCP server provides recommendation data only (not transactional booking) and requires a valid Expedia API key, which is typically obtained through an Expedia Group partner or account manager relationship.

Tools

Prerequisites

• Python 3.11 or higher (if running via uvx)
• A valid Expedia API key (obtained through Expedia Group partnership)
• EXPEDIA_API_KEY environment variable set

Option 1: Run via uvx (stdio)

export EXPEDIA_API_KEY="your_api_key_here"
uvx expedia_travel_recommendations_mcp --protocol "stdio"

Option 2: Run via Docker (streamable-http)

docker run \
  -p 9900:9900 \
  -e EXPEDIA_API_KEY=your_api_key_here \
  expediagroup/expedia-travel-recommendations-mcp

MCP Client Config (stdio)

{
  "mcpServers": {
    "expedia-travel-recommendations": {
      "command": "uvx",
      "args": ["expedia_travel_recommendations_mcp", "--protocol", "stdio"],
      "env": {
        "EXPEDIA_API_KEY": "your_api_key_here"
      }
    }
  }
}

MCP Client Config (streamable-http)

{
  "mcpServers": {
    "expedia-recommendation": {
      "url": "http://localhost:9900/mcp"
    }
  }
}

• Build a travel-planning AI assistant that recommends hotels filtered by amenities (pool, spa) and guest rating for a given destination and date range.
• Generate end-to-end trip itineraries by combining hotel, flight, activity, and car rental recommendations into a single AI workflow.
• Power a chatbot for a travel agency partner that surfaces Expedia's recommended properties and activities based on natural language queries.
• Compare hotel options across property types (HOTEL, RESORT, etc.) sorted by cheapest or highest-rated, then summarize results for a user.
• Suggest curated activities and tours alongside a booked hotel stay to upsell experiences in an agentic travel app.

• "Find hotels in Seattle from May 1 to May 5 with a pool and spa, sorted by cheapest."
• "Recommend resorts in Maui for July 10-17 with a wonderful guest rating."
• "Suggest top activities and tours in Tokyo for a 4-day trip."
• "Find a rental car at LAX for pickup on June 1 and return on June 5."
• "Plan a long weekend in Barcelona: pick a hotel, find flights from JFK, and suggest two activities."

• Official MCP server maintained by Expedia Group (ExpediaGroup GitHub org).
• Supports both stdio and streamable-http transports for flexible local or web deployment.
• Covers four major travel verticals (hotels, flights, activities, cars) through a single server.
• Distributed as both a Python package (uvx) and a Docker image for easy deployment.

• Requires an Expedia API key, which is gated behind Expedia Group's partner program rather than a self-serve signup.
• Provides recommendations only, not full booking/checkout transactions.
• Expedia's broader MCP program is described as in an "exploration phase," so capabilities and stability may evolve.

• Booking.com / hotel MCP community servers for alternative lodging data sources.
• Apify Expedia Hotels MCP (third-party scraper-based alternative when an official API key is unavailable).
• Direct Amadeus or Skyscanner API integrations for flight-centric travel agents.