Google Flights API & Scraper — Fares, Schedules & Price Insights

SEO title: Google Flights API — fares, schedules, price insights SEO description: Google Flights API for Python, JavaScript or no code: one-way and round-trip fares for any routes and dates, airlines, flight numbers, stops, CO2, plus Google's low/typical/high price verdict.

A Google Flights API you can call from Python, JavaScript, any HTTP client or an AI agent. Give it routes and dates and get every itinerary Google Flights shows, one row each:

• Fare: total for your passengers and cabin class, in your currency (round-trip total for round trips)
• Schedule: airlines, departure and arrival in local time, total duration, stops, layover airports and durations
• Every segment: flight number, aircraft, legroom, times
• CO2: Google's emission estimate vs. the typical flight on the route
• Price insight: Google's own verdict on the search ("prices are currently low / typical / high"), the typical price range, and optionally ~60 days of price history

Search many origins, destinations and dates in one run. Every combination is searched, so a list of dates gives you a fare calendar.

Why this exists

Google Flights has no public API, and the most-used scraper bills per page plus per booking option, so a single route lookup can cost several cents. This actor reads the same data the Google Flights results page is built from, without a browser:

• Searches are encoded the way the Google Flights app does it. Natural-language searches return no flights from cloud servers most of the time; encoded searches return flights.
• Fast: the first attempt goes out directly. Only when Google answers with an error does a retry go through a fresh proxy session.
• One flat row per itinerary, ready for a spreadsheet, a database or an AI agent.

Cloud test: 3 origins × 2 destinations × 5 dates = 30 searches → 300 itineraries in 43 seconds. Every search returned flights; 6 of 36 requests needed a retry.

Input

Field	What it does	Default
origins	IATA codes: airports (JFK) or metro areas (NYC, LON)	—
destinations	Same format	—
departureDates	YYYY-MM-DD list; each date is a search	—
tripLengthDays	Empty = one-way; e.g. 7 = round trip returning 7 days later	one-way
adults, children, infantsInSeat, infantsOnLap	Passengers (max 9)	1 adult
cabinClass	economy, premium_economy, business, first	economy
maxStops	any, nonstop, 1, 2	any
maxResultsPerSearch	Itineraries per search (Google lists ~10-40)	30
currency	USD, EUR, GBP...	USD
includePriceHistory	~60-day price history on each search's first row	false
language, countryCode	Interface language, point of sale	en-US, us
proxyConfiguration	Proxy used for retries	Apify Proxy

Worked example — input

{
  "origins": ["JFK"],
  "destinations": ["LHR"],
  "departureDates": ["2026-11-20"],
  "maxResultsPerSearch": 10
}

Worked example — output (one row, real data)

{
  "origin": "JFK",
  "destination": "LHR",
  "departureDate": "2026-11-20",
  "returnDate": null,
  "tripType": "one_way",
  "cabinClass": "economy",
  "passengers": {"adults": 1, "children": 0, "infantsInSeat": 0, "infantsOnLap": 0},
  "currency": "USD",
  "rank": 4,
  "category": "other",
  "price": 420,
  "airlines": ["Aer Lingus"],
  "departureAirport": "JFK",
  "departureTime": "2026-11-20T17:00",
  "arrivalAirport": "LHR",
  "arrivalTime": "2026-11-21T08:05",
  "durationMinutes": 605,
  "stops": 1,
  "layovers": [{"airport": "DUB", "airportName": "Dublin Airport", "city": "Dublin", "durationMinutes": 125}],
  "segments": [
    {"flightNumber": "EI 104", "airline": "Aer Lingus", "departureAirport": "JFK", "departureAirportName": "John F. Kennedy International Airport", "departureTime": "2026-11-20T17:00", "arrivalAirport": "DUB", "arrivalAirportName": "Dublin Airport", "arrivalTime": "2026-11-21T04:30", "durationMinutes": 390, "aircraft": "Airbus A330", "legroom": "31 inches"},
    {"flightNumber": "EI 152", "airline": "Aer Lingus", "departureAirport": "DUB", "departureAirportName": "Dublin Airport", "departureTime": "2026-11-21T06:35", "arrivalAirport": "LHR", "arrivalAirportName": "Heathrow Airport", "arrivalTime": "2026-11-21T08:05", "durationMinutes": 90, "aircraft": "Airbus A320neo", "legroom": "29 inches"}
  ],
  "emissions": {"grams": 394000, "typicalGrams": 432000, "differencePercent": -9},
  "priceInsights": {"priceLevel": "high", "priceLevelCode": 5, "lowestPrice": 420, "typicalPrice": 286, "typicalPriceRange": [170, 300]},
  "bookingToken": "CjRIOURJY3dQbEoxOTRB...",
  "googleFlightsUrl": "https://www.google.com/travel/flights/search?tfs=...",
  "hasData": true,
  "scrapedAt": "2026-10-12T15:13:02Z"
}

Output notes

• Times are local to each airport (2026-11-20T17:00 at JFK, 08:05 at LHR the next day), the same as on Google Flights.
• price is the total for all passengers. For round trips it is the round-trip total, and the row describes the outbound flight. Google only shows return-flight options after you pick an outbound flight, so they are not included.
• priceInsights describes the whole search and repeats on every row of it. priceLevel is Google's own verdict, typicalPriceRange is the range Google considers normal for the route. priceHistory (when requested) is only on rank 1.
• category: best is Google's "Best" list, other is everything else, ranked as Google ranks them.
• A search that fails or has no flights appears as one row with hasData: false and an error. That row is not charged.

Use it from code or an AI agent

curl -X POST "https://api.apify.com/v2/acts/westerly_breaker~google-flights-scraper/run-sync-get-dataset-items?token=$APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"origins": ["BUD"], "destinations": ["LON", "PAR"], "departureDates": ["2026-12-04", "2026-12-05"], "tripLengthDays": 3, "maxStops": "nonstop", "maxResultsPerSearch": 5}'

From Python (pip install apify-client):

from apify_client import ApifyClient

client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("westerly_breaker/google-flights-scraper").call(
    run_input={"origins": ["NYC"], "destinations": ["LON"], "departureDates": ["2026-11-20", "2026-11-21"], "maxResultsPerSearch": 5}
)
for flight in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(flight["departureDate"], flight["price"], flight["airlines"], flight["stops"])

Common uses: fare monitoring (schedule a daily run and watch price and priceLevel), finding the cheapest day in a date range, comparing routes and airlines for travel tools and agents, and route and schedule research.

Pricing (pay-per-event)

$0.001 per itinerary delivered, platform usage included. You control the cost with maxResultsPerSearch.

• 1 route × 14 dates × 10 itineraries = 140 × $0.001 = $0.14
• Searches that fail or have no flights are not charged.
• If you set a maximum cost per run, the actor stops cleanly at that limit. Every row you paid for is delivered, and you are never billed for a row you didn't get.

Error messages

Bad input stops the run before any request, naming the field and the fix, for example:

• 'origins' entry 'London' is not a 3-letter IATA code. Use airport codes like 'JFK', 'LHR' or metro-area codes like 'NYC', 'LON' (they cover all of the city's airports).
• Each infant on lap needs an adult: 'infantsOnLap' cannot exceed 'adults'.

Known limitations

• Return-flight details for round trips and per-site booking options (airline vs. agencies) each need extra requests per itinerary and are not included yet.
• Multi-city trips are not supported yet (one-way and round trip only).
• Fares change constantly and can differ by countryCode (point of sale), the same as on google.com.

Data source

Public Google Flights search results (google.com/travel/flights), the same data the web page shows. No login, no API key.