Claude Code

\n\n```bash\nclaude mcp add --transport http google-flights \"https://mcp.hasdata.com/api/mcp?apis=google_travel_flights\" \\\n --header \"x-api-key: HASDATA_API_KEY\"\n```\n\n

\n\n

Claude Desktop

\n\nSettings, then Connectors, then Add custom connector, then paste `https://mcp.hasdata.com/api/mcp?apis=google_travel_flights` and sign in.\n\nFor the config-file route, Claude Desktop loads only local (stdio) servers, so it reaches a remote server through a stdio launcher. The `@hasdata/google-flights-mcp` package is that launcher, and it reads the key from the environment. Add this to `claude_desktop_config.json`:\n\n```json\n{\n \"mcpServers\": {\n \"google-flights\": {\n \"command\": \"npx\",\n \"args\": [\"-y\", \"@hasdata/google-flights-mcp\"],\n \"env\": { \"HASDATA_API_KEY\": \"YOUR_KEY\" }\n }\n }\n}\n```\n\nFor Python instead of Node, swap the launcher for the PyPI package, which `uvx` runs without a manual install:\n\n```json\n{\n \"mcpServers\": {\n \"google-flights\": {\n \"command\": \"uvx\",\n \"args\": [\"hasdata-google-flights-mcp\"],\n \"env\": { \"HASDATA_API_KEY\": \"YOUR_KEY\" }\n }\n }\n}\n```\n\n

\n\n

Cursor

\n\n`~/.cursor/mcp.json` for every project, or `.cursor/mcp.json` for one:\n\n```json\n{\n \"mcpServers\": {\n \"google-flights\": {\n \"url\": \"https://mcp.hasdata.com/api/mcp?apis=google_travel_flights\",\n \"headers\": { \"x-api-key\": \"HASDATA_API_KEY\" }\n }\n }\n}\n```\n\n

\n\n

Windsurf

\n\n`~/.codeium/windsurf/mcp_config.json`. Windsurf calls the field `serverUrl`, not `url`:\n\n```json\n{\n \"mcpServers\": {\n \"google-flights\": {\n \"serverUrl\": \"https://mcp.hasdata.com/api/mcp?apis=google_travel_flights\",\n \"headers\": { \"x-api-key\": \"HASDATA_API_KEY\" }\n }\n }\n}\n```\n\n

\n\n

VS Code

\n\n`.vscode/mcp.json` in the workspace:\n\n```json\n{\n \"servers\": {\n \"google-flights\": {\n \"type\": \"http\",\n \"url\": \"https://mcp.hasdata.com/api/mcp?apis=google_travel_flights\",\n \"headers\": { \"x-api-key\": \"HASDATA_API_KEY\" }\n }\n }\n}\n```\n\n

\n\n## Example prompts\n\nPrompts, not code. Paste one in and the agent picks the tool itself. Each is annotated with the calls it takes, because every successful call costs 15 credits.\n\n> Find one-way flights from JFK to London Heathrow on September 15, sorted by price, and give me the three cheapest with airline and carbon estimate.\n\n*One call, 15 credits. Fares, legs and emissions all come back together.*\n\n> Same route but non-stop only, in business class, and tell me which option has the lowest emissions.\n\n*One call, 15 credits. Cabin and stops are filters on the one request.*\n\n> Is $295 a good price for JFK to LHR right now, given the price history?\n\n*One call, 15 credits. The response carries `priceInsights` with a typical range and a price level.*\n\n> Round trip JFK to LHR, out September 15 back September 22, cheapest fare.\n\n*Two calls, 30 credits. Google returns outbound options first, then the return leg is a second call keyed by the option you pick.*\n\nA round trip is two calls by design. The first returns outbound itineraries each with a `departureToken`, and you pass that token back to get the matching return flights. One-way and the price check are a single call each.\n\n## Tools\n\nOne tool, read-only. The sample below is trimmed from a real call, and fares move constantly. Read it as a shape. The tool name links to its endpoint reference, which carries the full parameter list.\n\nThe sample is the payload, not the whole response. A `tools/call` result carries one text block, and that text is itself JSON holding `url`, `status`, `text` and `json`, with the scraped data under `json`. From a raw JSON-RPC response the path is `result.content[0].text`, parsed, then `.json`. A chat client unwraps that for you and code talking to the endpoint directly does not.\n\n### Get Google Flights results\n\n[`hasdata_google_travel_flights_getGoogleFlights`](https://docs.hasdata.com/apis/google-travel/flights?utm_source=github&utm_medium=syndication&utm_campaign=google-flights-mcp)\n\nItineraries for a route and date, with fares, legs, emissions and price history.\n\n| Parameter | Type | Required | Notes |\n| :--- | :--- | :--- | :--- |\n| `departureId` | string | yes | IATA code such as `JFK`, or a location kgmid like `/m/02_286`. Comma-separate several airports |\n| `arrivalId` | string | yes | Same format as `departureId` |\n| `outboundDate` | string | yes | `YYYY-MM-DD` |\n| `type` | string | | `roundTrip` by default, `oneWay`, or `multiCity` with `multiCityJson` |\n| `returnDate` | string | | Required when `type` is `roundTrip` |\n| `travelClass` | string | | `economy`, `premiumEconomy`, `business` or `first` |\n| `stops` | string | | `nonStop`, `oneStopOrFewer` or `twoStopsOrFewer` |\n| `sortBy` | string | | `topFlights` default, plus `price`, `duration`, `emissions`, `departureTime`, `arrivalTime` |\n| `adults` / `children` / `infantsInSeat` / `infantsOnLap` | number | | Passenger mix |\n| `maxPrice` / `maxDuration` / `bags` | number | | Ceilings and carry-on count |\n| `includeAirlines` / `excludeAirlines` | string | | Comma-separated IATA airline codes, one or the other, not both |\n| `departureToken` | string | | Select an outbound option and fetch its return or next leg |\n| `bookingToken` | string | | Fetch booking options for a chosen itinerary |\n| `currency` / `gl` / `hl` | string | | Currency and the country and language of the search |\n| `deepSearch` | boolean | | Match what Google shows in a browser, slower to return |\n\nThe reference also documents `includeConnections`, `excludeConnections`, `layoverDuration`, `outboundTimes`, `returnTimes`, `showHidden`, `lessEmissions` and `multiCityJson`.\n\nResults split into `bestFlights` and `otherFlights`. Each itinerary carries `price`, `type`, `totalDuration` in minutes, a `flights` array of legs, a `carbonEmissions` object, and a `bookingToken`. Each leg holds the `departureAirport` and `arrivalAirport` (each with `id`, `name` and local `time`), `duration`, `airline`, `flightNumber`, `airplane`, `legroom`, `travelClass`, an `extensions` array, and `oftenDelayedByOver30Min` on legs Google flags. A non-stop itinerary has one leg, a connection has several.\n\n> `carbonEmissions` is in grams, not kilograms. `thisFlight: 433000` is 433 kg. `differencePercent` compares it to `typicalForThisRoute`, so a negative number is a greener-than-average flight.\n\n```json\n{\n \"price\": 295,\n \"type\": \"One way\",\n \"totalDuration\": 415,\n \"flights\": [\n {\n \"departureAirport\": { \"id\": \"JFK\", \"name\": \"John F. Kennedy International Airport\", \"time\": \"2026-09-15 8:15\" },\n \"arrivalAirport\": { \"id\": \"LHR\", \"name\": \"Heathrow Airport\", \"time\": \"2026-09-15 20:10\" },\n \"duration\": 415,\n \"airline\": \"Virgin Atlantic\",\n \"flightNumber\": \"VS 26\",\n \"airplane\": \"Boeing 787\",\n \"travelClass\": \"Economy\"\n }\n ],\n \"carbonEmissions\": { \"thisFlight\": 367000, \"typicalForThisRoute\": 419000, \"differencePercent\": -12 },\n \"bookingToken\": \"W1t7...\"\n}\n```\n\n`priceInsights` sits alongside the itineraries with `lowestPrice`, a `typicalPriceRange`, a `priceLevel` such as `typical`, and a `priceHistory` of `[timestamp, price]` points. `airports` echoes the resolved departure and arrival airports with city and country.\n\n## Errors and failure paths\n\nYour client almost never sees an HTTP error code from a tool call. The MCP layer answers 200 and puts the failure inside the result, with `isError` set to `true` and the reason as text. The agent reads a message where you might expect a status line.\n\n**A wrong key surfaces as tool output, not as a failed connection.** `tools/list` accepts any non-empty key and returns the tool, so the client completes its handshake and shows green. The first tool call then comes back with `isError: true` and the text `HasData API error: 401 Unauthorized`. Watch for that string, because nothing earlier in the flow reports the problem.\n\n**A missing key is the one real HTTP error.** Authorization runs before any tool, and the connection itself fails with 401. CORS headers are present, and a browser client reads the status and not an opaque network failure.\n\n**An argument that breaks the tool's schema is rejected before it becomes a scrape.** The server answers with `isError: true` and the text `MCP error -32602: Input validation error`, naming the offending field. A `roundTrip` without a `returnDate`, or `includeAirlines` together with `excludeAirlines`, is caught here.\n\n**A route with no flights on the date returns a successful result with the itinerary arrays empty**, not an error. `requestMetadata.status` still reads `ok`. Test for the flights before you rank them.\n\n**A bad airport code returns 400** with `requestMetadata.status` set to `error`. Use IATA codes or kgmids, not city names.\n\nResults that carry data also carry a `requestMetadata.id` worth quoting in support.\n\n## Pricing, free tier and limits\n\nEvery Google Flights call costs **15 credits per successful call**. Response size does not change the price, and the deep search costs the same as a standard one.\n\nThe free trial is **1,000 credits over 30 days with no card**, which is about 66 flight searches. After that an active account keeps getting 100 credits topped up each day whenever its balance drops below 100, so a low-volume agent runs on the free tier indefinitely.\n\nPaid plans start at **$49 a month** for 200,000 credits, which is about 13,000 searches. The unit price falls with volume, from **$3.68 per 1,000 calls** on the entry plan to **$1.49** on Business, **$1.25** on Growth and **$1.12** on the largest [high-volume plans](https://hasdata.com/prices?utm_source=github&utm_medium=syndication&utm_campaign=google-flights-mcp).\n\nYour plan also sets concurrency. The free trial allows 1 request at a time, Startup 15, Business 30, Growth 50, and the high-volume plans run from 200 to 1,500. Handle the overflow case defensively in anything unattended.\n\nA request that comes back non-200 is not billed. A round trip is two calls, so budget for it.\n\n## Tool selection\n\nThe `apis` query parameter decides which tools your agent sees. Fewer tools means less context spent on tool definitions, and fewer chances for the model to reach for the wrong one.\n\n```\n?apis=google_travel_flights the one tool in this repo\n?apis=google_travel add Google Hotels\n?apis=google_travel_flights,airbnb flights plus Airbnb stays\n```\n\nThe parameter takes provider names like `google_travel` and individual API names like `google_travel_flights`. Misspelled names are ignored. If every name is wrong the request fails with 400, and the body lists both what it did not recognise and every valid value. Drop the parameter and the same endpoint exposes all 57 HasData tools.\n\n## How it compares\n\nGoogle retired its QPX Express flight API in 2018 and never replaced it, so there is no official Google Flights API. The remaining routes are scraping the public results or licensing raw GDS fare data, which is heavy and expensive. This server reads the same results the site shows and returns them as JSON.\n\n| | Official Google API | This server |\n| :--- | :--- | :--- |\n| Availability | None since QPX Express closed in 2018 | Maintained schema over the live results |\n| Emissions data | Not offered | Per-itinerary, compared to the route average |\n| Price history | Not offered | `priceInsights` with a typical range |\n| Setup | Nothing to set up, because it does not exist | One key and one URL |\n| Cost | Not applicable | Paid past the trial, 15 credits a call |\n\n**What this server does not do.** No booking and no payment. It reads fares, legs and the tokens Google itself uses to move to booking, and hands the booking step back to you.\n\n## FAQ\n\n### Is there an official Google Flights API?\n\nNo. Google closed QPX Express in 2018 and has not shipped a replacement. Every option reads the same public results the website serves. This one is maintained by HasData and returns them as structured JSON.\n\n### What is a Google Flights MCP server?\n\nA server that exposes Google Flights as a tool an AI client can call. The client sends a tool call over the Model Context Protocol, the server fetches the itineraries and returns structured JSON, and the model works with the result. This one exposes a single tool and runs remotely.\n\n### Why is a round trip two calls?\n\nGoogle returns outbound options first, each with a `departureToken`. You pick one and pass its token back to get the return flights that pair with it. That mirrors how the site works, and it is why a round trip costs 30 credits.\n\n### Are the carbon numbers in kilograms?\n\nNo, grams. `thisFlight: 433000` means 433 kg, and `differencePercent` compares it to the route average.\n\n### What is deep search?\n\nA slower mode that returns exactly what Google Flights shows in a browser. Leave it off for speed, turn it on when you need parity with the site.\n\n### Can I use this together with other HasData APIs?\n\nYes. The `apis` parameter takes a list, and `?apis=google_travel` adds Google Hotels alongside flights. [Drop the parameter](#tool-selection) and you get everything.\n\n### Compliance and personal data\n\nHasData accesses publicly available data only. A platform's terms may restrict automated access, and you are responsible for your own compliance.\n\n## HasData links\n\n| | |\n| :--- | :--- |\n| Product page and request builder | [Google Flights API](https://hasdata.com/apis/google-flights-api?utm_source=github&utm_medium=syndication&utm_campaign=google-flights-mcp) |\n| Server documentation | [MCP server docs](https://docs.hasdata.com/mcp-server?utm_source=github&utm_medium=syndication&utm_campaign=google-flights-mcp) |\n| All 57 tools in one server | [HasData/hasdata-mcp](https://github.com/HasData/hasdata-mcp) |\n| Client walkthroughs | [MCP clients and integrations](https://hasdata.com/integrations/mcp?utm_source=github&utm_medium=syndication&utm_campaign=google-flights-mcp) |\n| Everything else we scrape | [Google Flights API and 54 more](https://hasdata.com/apis/?utm_source=github&utm_medium=syndication&utm_campaign=google-flights-mcp) |\n| Plans and credit costs | [Plans and credit costs](https://hasdata.com/prices?utm_source=github&utm_medium=syndication&utm_campaign=google-flights-mcp) |\n| Keys and usage | [HasData dashboard](https://app.hasdata.com?utm_source=github&utm_medium=syndication&utm_campaign=google-flights-mcp) |\n| Node launcher on npm | [@hasdata/google-flights-mcp](https://www.npmjs.com/package/@hasdata/google-flights-mcp) |\n| Python launcher on PyPI | [hasdata-google-flights-mcp](https://pypi.org/project/hasdata-google-flights-mcp/) |\n\n## Development\n\nThis repository is configuration and documentation for a remote server. There is no build step and nothing to containerize.\n\nThe tests in `test/` assert the tool contract, the part that can break without a commit here. They check that `?apis=google_travel_flights` returns exactly one tool, that it still declares its required parameters, that the name has not changed, and that the key in use is actually accepted. That last check calls the tool for real and costs 15 credits, which is the price of a canary that can fail for the right reason.\n\n```bash\n# macOS and Linux\nHASDATA_API_KEY=your_key_here npm test\n\n# Windows PowerShell\n$env:HASDATA_API_KEY=\"your_key_here\"; npm test\n```\n\nThe same suite runs in CI on every push and once a week on a schedule, because the upstream tool list can change without anyone touching this repository. A failure means the tool list moved, the key stopped working, or the endpoint was unreachable, and the assertion message says which.\n\n## Contributing\n\nCorrections to the parameter table and the response sample are the most useful contribution, because those are the parts that drift. Include the call you made and the response you got. Pull requests from forks run the suite without a key, and the live checks skip instead of going red.\n\n## License\n\nMIT. See [LICENSE](https://github.com/HasData/google-flights-mcp/blob/main/LICENSE).\n","changelogMarkdown","videoKey","requiredSecretsJson","[{\"name\":\"x-api-key\",\"scope\":\"per-user\",\"header\":\"x-api-key\"}]","publishedAt","priceCents","trialDays",0,"upstream",{"_116":117,"_118":119,"_120":121,"_122":123,"_124":125,"_126":127},"requiredSecrets",[110],"tags",[99,100,101,102,103,104,19,105,106,107,108,109],"assets",[],"stats",{"_97":86,"_98":86},"ratingAvg","reviewCount","ai-agents","airfare","claude","cursor","flights","google-flights","itineraries","llm-tools","travel","web-scraping","windsurf",{"_111":112,"_113":114,"_115":112},"name","x-api-key","scope","per-user","header","topics",[99,100,101,102,103,104,19,105,106,136,137,138,107,108,109],"license","MIT","language","JavaScript","registryName","com.hasdata/google-flights","homepage","https://hasdata.com/apis/google-flights-api","packages",[128,129],{"_130":135,"_132":22,"_75":134},{"_130":131,"_132":133,"_75":134},"registryType","pypi","identifier","hasdata-google-flights-mcp","1.0.2","npm","mcp","mcp-server","model-context-protocol","ENV",{"_141":142,"_143":144,"_145":146},"ENVIRONMENT","production","TURNSTILE_SITE_KEY","0x4AAAAAAB8-rEI2gNDf_QhU","SENTRY_DSN","https://45fd1229f78089d5c3ee78be128030e1@o4510583261757440.ingest.de.sentry.io/4510584253448272"]

HasData Google Flights