Google Flights Search Scraper

Give it a route and a date and it hands back the whole fare table Google Flights shows for that search — every flight, with its fare, airlines, flight numbers, aircraft, departure and arrival times, stop count, layover airports and layover lengths, cabin, legroom and carbon figure. One-way or return, any cabin, up to nine travellers, priced for the market you sell into. Search one route on one day, or dozens of routes across dozens of dates in a single run. No account and no login are needed — just run it.

What you can do with it

• Track a route's price — run it daily on the routes you care about and keep a clean history of the cheapest fare, the airline holding it and the flight behind it.
• Watch competitors — see who is flying a route, at what times, on what aircraft, and where their fares sit against yours.
• Fill a travel site or app — a ready-made fare table per route and date, with airlines, times, stops and layovers already separated into columns.
• Plan a trip programme — compare thirty departure days at once and see how the fare moves through the month.
• Study a market — the same route in the United States, the United Kingdom, Germany or India, each priced the way travellers there see it.
• Report on emissions — every flight carries its own carbon figure, the typical figure for that route, and how far above or below typical it sits.

What you get

One row per flight. Abridged real example:

{
  "searchId": "JFK-LAX-2026-09-15",
  "source": "google-flights",
  "market": "US",
  "collectedAt": "2026-09-01T09:14:22.187Z",
  "origin": "JFK",
  "destination": "LAX",
  "departureDate": "2026-09-15",
  "returnDate": null,
  "tripType": "one_way",
  "cabinClass": "economy",
  "passengers": { "adults": 1, "children": 0, "infantsInSeat": 0, "infantsOnLap": 0 },
  "price": 169,
  "priceExact": 168.4,
  "isPriceExact": false,
  "currency": "USD",
  "currencyFromSource": true,
  "priceCovers": "one_way",
  "isCheapestInSearch": true,
  "rank": 1,
  "listing": "best",
  "itineraryId": "F0bbN",
  "departureAirport": "JFK",
  "departureAirportName": "John F. Kennedy International Airport",
  "arrivalAirport": "LAX",
  "arrivalAirportName": "Los Angeles International Airport",
  "departureLocal": "2026-09-15T08:00",
  "arrivalLocal": "2026-09-15T10:59",
  "totalDurationMin": 359,
  "stops": 0,
  "airlineCodes": ["DL"],
  "airlineNames": ["Delta"],
  "flightNumbers": ["DL713"],
  "aircraftTypes": ["Boeing 767"],
  "layoverAirports": [],
  "layoverDurationsMin": [],
  "legroom": "31 in",
  "co2eKg": 241,
  "routeTypicalCo2eKg": 337,
  "co2ePercentVsTypical": -28,
  "legs": [
    {
      "from": "JFK",
      "to": "LAX",
      "airlineCode": "DL",
      "airlineName": "Delta",
      "flightNumber": "DL713",
      "aircraft": "Boeing 767",
      "cabin": "economy",
      "legroom": "31 in",
      "departureLocal": "2026-09-15T08:00",
      "arrivalLocal": "2026-09-15T10:59",
      "durationMin": 359,
      "co2eKg": 241.3
    }
  ]
}

Alongside the table, each run writes a summary that separates flights collected, searches that returned flights, searches that genuinely had none, and searches that could not be completed. Those three are never mixed together.

Input reference

Field	Type	Default	What it does
origins	list of text	["JFK"]	Where the trip starts. Three-letter airport codes (JFK, LHR) or city codes covering every airport in a city (NYC, LON, PAR).
destinations	list of text	["LAX"]	Where the trip ends, in the same format. Every departure point is searched against every arrival point.
departureDates	list of text	["+30"]	Outbound dates. An exact day (2026-09-15) or a number of days ahead (+30), which keeps a scheduled run looking the same distance into the future every time.
tripLengthDays	number	empty	Empty means one-way. A number of nights turns the search into a return trip: leaving on the 15th with 7 nights comes back on the 22nd.
cabinClass	text	economy	economy, premium_economy, business or first.
stops	text	any	any, nonstop, one_stop_or_fewer or two_stops_or_fewer.
adults	number	1	Adults (12+). Nine travellers in total at most.
children	number	0	Children aged 2–11.
infantsInSeat	number	0	Infants in their own seat. See the limits below before using it.
infantsOnLap	number	0	Infants on an adult's lap; one adult each. See the limits below before using it.
market	text	US	The country you are selling into. Fares and availability differ by market.
currency	text	USD	The currency to ask for. Every row also states the currency that actually came back.
maxResultsPerSearch	number	100	How many flights to keep per route and date, cheapest first.
maxSearches	number	100	Safety cap on the total number of route-and-date combinations in one run.
cheapestOnly	true/false	false	Keep only the cheapest flight of each search — one row per route and date.
maxPrice	number	empty	Drop flights above this fare, in the currency that comes back.
maxJourneyMinutes	number	empty	Ask for flights no longer than this, layovers included.
includeAirlines	list of text	empty	Keep only flights on these airlines (BA, AA, LH).
parallelSearches	number	4	How many searches to run at the same time, 1 to 6.

Output fields

Field	What it is
searchId	The route and date this row answers, e.g. JFK-LAX-2026-09-15.
origin, destination	What you asked for (an airport or a city code).
departureDate, returnDate	The dates searched. returnDate is empty for one-way.
tripType	one_way or round_trip.
cabinClass, passengers	The cabin and traveller mix the fare applies to.
price	The fare as the source lists it, in whole units.
priceExact	The fare to the cent where the source publishes it, otherwise empty.
isPriceExact	Whether the listed fare is the exact one.
currency	The currency the fare actually came back in — read this, not the one you asked for.
currencyFromSource	True when the source stated the currency itself.
priceCovers	one_way, or round_trip_total when the fare covers the whole return trip.
isCheapestInSearch	True when this fare is the lowest found for that route and date.
rank	Position within the search, cheapest first.
listing	best when the source picked this flight out, other for the rest of the list.
itineraryId	The source's own identifier for this flight combination.
departureAirport, arrivalAirport	The actual airports flown, with their full names.
departureLocal, arrivalLocal	Local clock times at each airport.
totalDurationMin	Door-to-door journey length in minutes.
stops	Number of stops.
airlineCodes, airlineNames	The airlines on the itinerary.
flightNumbers	Every flight number, in order.
aircraftTypes	Aircraft flown.
layoverAirports, layoverDurationsMin	Where the flight stops and for how long.
legroom	Legroom on the first flight, as published.
co2eKg	Carbon for this flight.
routeTypicalCo2eKg	Carbon for a typical flight on this route — a figure about the route, not about this flight, and the same on every row of a search.
co2ePercentVsTypical	How this flight compares with that typical figure, in whole percent. -28 means 28% cleaner than typical; a positive number means dirtier. Published by the source, not calculated here.
legs	Each flight in the itinerary: airports, airline, flight number, aircraft, cabin, legroom, times, length and carbon.
bookingToken	The source's own reference for this itinerary.
market, collectedAt, source	Which market the fare was collected for, and when.

Pricing

You pay for what you get, in two parts:

• $0.003 per flight added to your results — $3.00 per 1,000 flights.
• $0.015 per route and date searched, charged once each search is complete. A route that genuinely has no flights on the day you searched is still a completed search and is charged. A search that could not be completed is not charged.

Worked example. Ten routes across seven dates is 70 searches — $1.05. If those routes return about 25 flights each, that is roughly 1,750 flights — $5.25. Total for the run: about $6.30.

Tracking prices instead? Turn on cheapestOnly and the same 70 searches produce 70 rows: $1.05 for the searches plus $0.21 for the rows, about $1.26.

Limits & what this actor cannot do

• Fares are a snapshot. Airline pricing changes continuously; a fare seen now may be gone minutes later.
• It cannot book, hold or reserve anything. It reports what the source publishes and cannot guarantee a fare is still available.
• The fare is the one shown for the search you asked for. Taxes, carrier fees and baggage rules vary by airline and fare type, and extras appear only where the source publishes them.
• The listed fare is rounded up to a whole unit (169 where the true fare is 168.40). Where the exact figure is published it is given in priceExact.
• Fares and availability differ by market and currency. Results reflect the market you choose, and each row carries the currency that actually came back — which is not always the one you asked for, and is never assumed.
• Dates run to 330 days ahead. A date in the past or beyond that window is reported as out of range rather than searched, because the source answers such a date with a normal-looking page that has no fares on it.
• Routes and schedules are seasonal. A real route with nothing flying on the dates you searched is reported as having no flights, not as an error.
• A search the source refuses to complete is reported as a failed search, never as "no flights found". Failed searches are listed individually in the run summary, and a run where nothing could be completed ends as a failed run rather than as an empty success.
• Return trips list the outbound flights, and the fare on each row is the full return fare, marked as round_trip_total.
• Trips with three or more separate legs are not supported. Search them as separate one-way legs instead.
• Searches that include an infant come back with no flights, whatever the route — the source does not publish results for them. The run warns you rather than leaving you to read an empty result as "nothing flies this route".
• The airline and maximum-fare filters are applied to the flights that come back, not to the search itself, so they narrow your results rather than making the search cheaper.
• The route's carbon figure (routeTypicalCo2eKg) describes the route, not the individual flight; it is the same on every row of a search by design, and it is a typical flight, not the cleanest one. Plenty of flights come in below it — that is what co2ePercentVsTypical being negative means, and it is normal rather than a sign of a bad figure.
• There is no "lowest carbon on this route" column, because the source does not publish one. Comparing a flight against the cleanest option on its route can only be done across the flights the run actually collected.
• The source publishes a limited number of flights for any one search, so a very broad search returns what it makes available rather than an unlimited list. Vary the dates to widen coverage.
• Speed depends on the size of the job and on the source's own response times; no fixed rate is promised.
• The source's terms govern automated access. You are responsible for using the data lawfully and in line with those terms.

FAQ

Do I need an account on Google Flights? No. Nothing is signed in and nothing is stored on your behalf.

Does it need my login or password? No. There is no credential of any kind to supply.

Can I schedule it? Yes. Schedule it like any other Actor. Writing dates as +30 rather than as a fixed day keeps a daily run looking the same distance ahead every morning, which is what you want for price tracking.

Are these fares bookable? They are the fares published for that search at the moment it was collected. Airline pricing moves constantly, and nothing here holds a seat, so treat a fare as an observation rather than a guarantee.

A search came back with no flights — is that a mistake? No. Routes are seasonal, and some airport pairs simply have nothing flying on a given day. That is reported as a completed search with no flights, and it is kept strictly separate from a search that could not be completed — you will never be shown a refusal dressed up as "no flights found".

Can I cover a whole month? Yes. Put up to sixty dates in departureDates (or a range such as +1, +2, +3 …) and each one becomes its own search, subject to maxSearches.

Why do the same route and date give different fares in different markets? Because they genuinely differ. Airlines file different fares by point of sale, so pick the market you sell into and compare within it rather than converting between markets.