Amadeus MCP Server
Enables flight search capabilities through the Amadeus Flight Offers Search API, allowing users to search for flights between locations with specified dates using natural language queries.
README
Amadeus MCP Server
MCP-Amadeus is a community-developed Model Context Protocol (MCP) server that integrates with the Amadeus Flight Offers Search API to provide flight search capabilities through natural language interfaces. Built for use with MCP-compatible clients (e.g., Claude Desktop).
This project enables users to easily search for flight options between two locations with specific dates using the power of large language models (LLMs) and the Amadeus API.
This project uses the official amadeus-python SDK
Disclaimer: This is an open-source project not affiliated with or endorsed by Amadeus IT Group. Amadeus® is a registered trademark of Amadeus IT Group.
✨ Features
✈️ Flight Offers Search
Retrieve flight options between two locations for specified dates.
"I'm looking for nonstop flights from New York to London on June 15th, any airline, for 1 adult."
→ ✈️ Returns available flight options with details like departure time, arrival time, airline, and price.
- Powered by Amadeus Flight Offers Search API
- Configurable test/production endpoints
- Requires origin, destination, number of tickets and travel date input
🌐 Demo
Once installed and connected to an MCP-compatible client (e.g., Claude Desktop), this server exposes tools that your AI assistant can use to fetch flight data.
🚀 Quick Start
Installing via Smithery
To install Amadeus MCP Server for Claude Desktop automatically via Smithery:
npx -y @smithery/cli install @donghyun-chae/mcp-amadeus --client claude
1. Clone and Setup
git clone https://github.com/donghyun-chae/mcp-amadeus.git
cd mcp-amadeus-flight-offers
# Install dependencies (using uv or pip)
uv sync
2. Get Your API Key and Set Environment
cp .env.example .env
Then edit .env and add your API credentials:
AMADEUS_CLIENT_ID=your_client_id
AMADEUS_CLIENT_SECRET=your_client_secret
AMADEUS_HOSTNAME=test
Configuration Options:
AMADEUS_CLIENT_ID: Your Amadeus API Client IDAMADEUS_CLIENT_SECRET: Your Amadeus API Client SecretAMADEUS_HOSTNAME: API endpoint to use -testfor testing (default) orproductionfor live API
Sign up on https://developers.amadeus.com/ and create an app to obtain your Client ID and Client Secret.
🌐 Environment Configuration
The server supports both test and production Amadeus API endpoints:
-
Test Environment (
AMADEUS_HOSTNAME=test): Use this for development and testing. The test environment provides:- Free API access with test data
- No charges for API requests
- Limited to test flight data
-
Production Environment (
AMADEUS_HOSTNAME=production): Use this for live applications. The production environment provides:- Real, live flight data
- Pay-per-request pricing
- Full access to all Amadeus services
⚠️ Important: Always use the test environment during development. Switch to production only when your application is ready for live data and you understand the associated costs.
3. Configure MCP Client
Register this server in your MCP client (e.g., Claude for Desktop).
Edit ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"amadeus": {
"command": "/ABSOLUTE/PATH/TO/PARENT/FOLDER/uv",
"args": [
"--directory",
"/ABSOLUTE/PATH/TO/PARENT/FOLDER/src/",
"run",
"--env-file",
"/ABSOLUTE/PATH/TO/PARENT/FOLDER/.env",
"server.py"
]
}
}
}
Replace
/ABSOLUTE/PATH/TO/PARENT/FOLDER/with the actual path to your project folder.
my case:
{
"mcpServers": {
"amadeus": {
"command": "/Users/asena/.local/bin/uv",
"args": [
"--directory",
"/Users/asena/mcp-amadeus/src/",
"run",
"--env-file",
"/Users/asena/mcp-amadeus/.env",
"server.py"
]
}
}
}
🛠️ Tools
After installation, the following tool is exposed to MCP clients:
search_flight_offers
Retrieves flight offers from the Amadeus Flight Offers Search API. The API endpoint (test/production) is configurable via the AMADEUS_HOSTNAME environment variable.
Request:
{
"action": "tool",
"name": "search_flight_offers",
"params": {
"originLocationCode": "JFK",
"destinationLocationCode": "LHR",
"departureDate": "2025-06-15",
"adults": 1
}
}
Parameters:
| Name | Type | Required | Description | Example |
|---|---|---|---|---|
| originLocationCode | string | Yes | IATA code of departure city/airport | JFK |
| destinationLocationCode | string | Yes | IATA code of destination city/airport | LHR |
| departureDate | string | Yes | Departure date (YYYY-MM-DD) | 2025-06-15 |
| adults | integer | Yes | Number of adults (1-9). Default: 1 | 2 |
| returnDate | string | No | Return date (YYYY-MM-DD). One-way if omitted | 2025-06-20 |
| children | integer | No | Number of children (2-11). Max total: 9 | 1 |
| infants | integer | No | Number of infants (≤2). Max: # of adults | 1 |
| travelClass | string | No | Cabin class: ECONOMY, BUSINESS, etc. | ECONOMY |
| nonStop | boolean | No | If true, only non-stop flights. Default: false | true |
| currencyCode | string | No | Currency in ISO 4217 (e.g., USD) | EUR |
| maxPrice | integer | No | Max price per traveler | 500 |
| max | integer | No | Max number of offers. Default: 250 | 10 |
Output: Returns flight offers in JSON format with airline, times, duration, and pricing details from Amadeus.
📚 References
📝 License
MIT License © 2025 donghyun-chae
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。