Access Real-Time Aviation Data

An MCP server exposing the AviationStack API as 12 tools covering flight, aircraft, and reference data. Flight tools: flights_with_airline returns a random sample of flights for a named airline; historical_flights_by_date fetches historical flights for a given date with optional airline/departure/arrival IATA filters (requires a Basic plan or higher); flight_arrival_departure_schedule returns arrival or departure schedules for a specific airport and airline; and future_flights_arrival_departure_schedule returns future scheduled flights for an airport, airline, and date. Random-sample lookup tools: random_aircraft_type, random_airplanes_detailed_info, random_countries_detailed_info, and random_cities_detailed_info each return a requested count of random records. Reference-data listing tools: list_airports, list_airlines, list_routes, and list_taxes all support limit/offset pagination, and list_airports/list_airlines/list_taxes additionally support a free-text search filter while list_routes filters by airline/departure/arrival IATA codes rather than free text.

Use it when an AI assistant needs live or historical flight schedules, aircraft and airplane details, or airport/airline/route/tax reference data - for example, checking a specific airline's current flights, an airport's arrival board, or looking up airport codes by search term. It requires a free or paid AviationStack API key (a free tier is available at signup) and Python 3.13 or newer with the uv package manager. Historical flight lookups specifically require a Basic AviationStack plan or higher - the free tier won't support that tool.

12 tools spanning live/historical/future flight schedules by airline, airport, or date; random-sample lookups for aircraft types, airplanes, countries, and cities; and paginated, searchable listings of airports, airlines, routes, and taxes - all implemented as Python functions decorated with @mcp.tool() on top of the FastMCP class from mcp.server.fastmcp.

{
  "mcpServers": {
    "Aviationstack MCP": {
      "command": "uvx",
      "args": ["aviationstack-mcp"],
      "env": { "AVIATION_STACK_API_KEY": "<your-api-key>" }
    }
  }
}

That's the recommended zero-clone install via uvx. Alternatively, clone the repository and run it locally by pointing uv's --directory at the cloned src/aviationstack_mcp path and running -m aviationstack_mcp mcp run. Either way, an AVIATION_STACK_API_KEY environment variable is required.

Developers building travel, transportation, or aviation-data applications who want live/historical flight schedules and aviation reference data available through an MCP-compatible AI assistant instead of calling the AviationStack REST API directly - listed in the official modelcontextprotocol/servers community directory as well as Smithery and Glama. Also listed on mcpservers.org and in punkpeye/awesome-mcp-servers under travel and transportation. Distributed under the MIT License, with the main server logic living in server.py.