title: Google Flights MCP emoji: ✈️ colorFrom: blue colorTo: indigo sdk: docker pinned: false

Google Flights MCP Server

A Model Context Protocol (MCP) server for intelligent flight search and travel planning

Overview

Integrates Google Flights data directly into your AI workflow with natural language searches, intelligent price comparisons, and automated travel planning.

Data Sources:

• fast-flights v3 (Primary): Free, no API key required — JS parser, connection segments, native airline filters; round-trip searches return outbound options with RT total prices
• SerpAPI (Fallback): Richer data including flight numbers, layovers, carbon emissions, and multi-seller booking options — used when SERPAPI_API_KEY is set (250 free searches/month)

Tools

Resources

• airports://all — Notes on IATA usage (v3 no longer ships an airport enum)
• airports://{code} — Validate a 3-letter IATA code (e.g. airports://LAX)

Prompts

11 built-in travel planning prompts: find_best_deal, weekend_getaway, last_minute_travel, business_trip, family_vacation, budget_backpacker, loyalty_program_optimizer, holiday_peak_travel, long_haul_international, stopover_explorer, reliable_search_strategy.

Quick Start

Option 1: Install from PyPI

uvx mcp-server-google-flights

Option 2: Run from source

git clone https://github.com/HaroldLeo/google-flights-mcp.git
cd google-flights-mcp
python -m venv .venv
source .venv/bin/activate
pip install -e .

Configuration

Claude Desktop

{
  "mcpServers": {
    "google-flights": {
      "command": "uvx",
      "args": ["mcp-server-google-flights"],
      "env": {
        "SERPAPI_API_KEY": "your_serpapi_key_here"
      }
    }
  }
}

From source

{
  "mcpServers": {
    "google-flights": {
      "command": "/absolute/path/to/.venv/bin/python",
      "args": ["/absolute/path/to/google-flights-mcp/src/mcp_server_google_flights/server.py"],
      "env": {
        "SERPAPI_API_KEY": "your_serpapi_key_here"
      }
    }
  }
}

Get a free SerpAPI key: serpapi.com/users/sign_up (250 free searches/month)

Remote Deployment (Hugging Face Spaces)

This server supports remote deployment as an SSE MCP server.

Deploy to HF Spaces

1. Create a new Space at huggingface.co/new-space with Docker SDK
2. Push this repository to the Space
3. (Optional) Add SERPAPI_API_KEY as a Space secret

The Dockerfile and transport switching are already configured — set MCP_TRANSPORT=sse (done automatically in the Dockerfile) to run in SSE mode.

Connect to a deployed Space

{
  "mcpServers": {
    "google-flights": {
      "url": "https://YOUR-USERNAME-google-flights-mcp.hf.space/sse"
    }
  }
}

Environment Variables

Troubleshooting

No flights found: Try a different date or route. Google Flights rate-limits scraping — SerpAPI fallback is more reliable.

Scraping errors: Retry later, or set SERPAPI_API_KEY for automatic fallback on search_one_way_flights and search_round_trip_flights. Date-range searches do not use SerpApi. search_flights_by_airline uses the SerpApi fallback for one-way searches with airline codes outside the built-in name table when SERPAPI_API_KEY is set.

Slow searches: search_round_trips_in_date_range is limited to 30 date combinations to avoid rate limiting. Narrow your date range or use min_stay_days/max_stay_days.

License

MIT — see LICENSE