About Sparkflights
Sparkflights is a hosted MCP server for finding verified flight fares by route, month, cabin, region, and price. It is for travelers and MCP clients seeking fare details and booking links; results reflect fares Sparkflights has verified, not a live search.
Key features
- 01Search verified fares by route and month
- 02Filter by cabin, region, and price
- 03Rank results by cost per mile
- 04Retrieve itinerary, flights, and date prices
- 05Open fare and route pages
Use cases
- 01Find fares between cities for a chosen month
- 02Search for cabin-specific fares below a price limit
- 03Inspect a result's flights and listed dates
MCP support
Sparkflights exposes two read-only tools through a hosted Streamable HTTP MCP server: `search_deals` and `get_deal`. MCP clients can search verified fares and retrieve details, with results linking to fare pages for booking.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://www.sparkflights.com/mcp/searchclaude mcp add --transport http sparkflights-com-sparkflights https://www.sparkflights.com/mcp/search{"mcpServers":{"sparkflights-com-sparkflights":{"type":"http","url":"https://www.sparkflights.com/mcp/search"}}}Open in Cursor{"servers":{"sparkflights-com-sparkflights":{"type":"http","url":"https://www.sparkflights.com/mcp/search"}}}Open in VS Code{"mcpServers":{"sparkflights-com-sparkflights":{"type":"http","url":"https://www.sparkflights.com/mcp/search"}}}{"mcpServers":{"sparkflights-com-sparkflights":{"serverUrl":"https://www.sparkflights.com/mcp/search"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
Details
From the project READMESparkflights MCP server
Search flight fares that Sparkflights has found and verified, from Claude, ChatGPT, Cursor, VS Code or any other MCP client.
https://www.sparkflights.com/mcp/search
The server is hosted. There is nothing to install, no account and no API key. It uses streamable HTTP.

Tools
Both tools only read data.
search_dealsfinds verified fares by origin and destination (an airport code such asJFK, or a city such asNew York), departure month, cabin, region and maximum price. Results are ranked by cost per mile.get_dealreturns one fare's itinerary, flights and the date pairs its fare page lists, with their prices.
Each result links to its Sparkflights fare page, where you book it, and to its route page, where you can follow the route for new fares.
Sparkflights is not a live fare search. An empty result means no verified fare matches right now, not that no flights exist. Sparkflights does not sell tickets or take payments.
Add it to your client
Claude (claude.ai or Claude Desktop): Settings → Connectors → Add custom connector, then paste the URL above.
Claude Code:
claude mcp add --transport http sparkflights https://www.sparkflights.com/mcp/search
Cursor (~/.cursor/mcp.json):
{
"mcpServers": {
"sparkflights": {
"url": "https://www.sparkflights.com/mcp/search"
}
}
}
VS Code (.vscode/mcp.json):
{
"servers": {
"sparkflights": {
"type": "http",
"url": "https://www.sparkflights.com/mcp/search"
}
}
}
ChatGPT: turn on developer mode, then add a custom MCP server with the URL above.
Other clients: mcp.json has the standard mcpServers entry.
Try asking
- Find flight deals from New York to London
- Are there any business class flight deals from San Francisco?
- Flights from Chicago in March under $500
- Any flight deals to Tokyo?
- Find flight deals from Los Angeles, then show me the dates and flights for the first one
Links
- Sparkflights
- For agents: how AI agents submit the fares they find and get credited for them
- Privacy · Terms
- Problems with the server: open an issue here, or email [email protected]
Publishing
server.json is published to the MCP Registry by .github/workflows/publish-mcp.yml when a v* tag is pushed. Bump version in server.json before tagging.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.