Skyscanner Flights Scraper

Skyscanner Flights Scraper: fares for the routes and dates you pick, one-way, return or multi-city

Give it routes and dates, or paste Skyscanner links. Each search comes back as one row per priced itinerary: the lowest price, who is selling it, the airlines, the flight numbers, local departure and arrival times, journey length and stops.

The thing to plan around is the wait. A Skyscanner search takes about three minutes to settle, because that is how long the site itself takes to collect quotes. Searches in a run wait together, so five routes take about as long as one, but no single search comes back quickly.

✈️ What Skyscanner Flights Scraper does

Three trip shapes, in one field. A one-way is {"origin": "JFK", "destination": "LAX", "date": "2026-10-20"}, or the short form "JFK-LAX-2026-10-20". Add returnDate and it is a return. Put two or three legs under legs and it is a multi-city trip.

Add lastDate to a one-way or return and the run searches every departure day from date to that day, seven at most. Each day is a search of its own, and on a return the return date moves along with the departure.

Airport codes like LHR work, and so do city codes like LON, NYC and PAR, which cover every airport in that city. Codes are checked before anything runs, so a typo is refused rather than quietly searched.

Filters and order run on the full list a search returned, before anything is delivered, so sortBy: "fastest" with maxItinerariesPerRoute: 10 gives the ten quickest that pass.

📥 What you give it

{
  "routes": [
    { "origin": "JFK", "destination": "LAX", "date": "2026-10-20", "returnDate": "2026-10-27" },
    "LHR-BCN-2026-11-03"
  ],
  "market": "UK",
  "currency": "GBP",
  "maxStops": "1",
  "sortBy": "cheapest",
  "maxItinerariesPerRoute": 50
}

best is this actor's own ranking, not the one on Skyscanner's site. It is each itinerary's price over the cheapest price, plus its journey time over the quickest, lowest first. If you need Skyscanner's own idea of "best", this is not it.

Dates run from yesterday to 365 days out. A busy route has several hundred itineraries and a return over a thousand, so set maxItinerariesPerRoute deliberately: you pay for each one delivered.

📤 What you get back

A real row, from run a7USyKNkEZY5JHS8d, with the middle of its three flights cut out here:

{
  "ok": true, "recordType": "itinerary",
  "route": "JFK-MIA", "tripType": "one-way",
  "date": "2026-10-24", "returnDate": null,
  "price": 171, "currency": "USD", "pricingOptions": 2, "cheapestSeller": "Kiwi.com",
  "carriers": ["jetBlue", "Frontier Airlines"],
  "flightNumbers": ["B61201", "B61798", "F93442"],
  "departureAirport": "JFK", "arrivalAirport": "MIA",
  "departure": "2026-10-24T15:59:00", "arrival": "2026-10-25T10:15:00",
  "durationMinutes": 1096, "stops": 2,
  "legs": [
    { "from": "JFK", "to": "MIA",
      "departure": "2026-10-24T15:59:00", "arrival": "2026-10-25T10:15:00",
      "durationMinutes": 1096, "stops": 2,
      "carriers": ["jetBlue", "Frontier Airlines"],
      "flights": [
        { "flightNumber": "B61201", "carrier": "jetBlue", "carrierCode": "B6",
          "operatedBy": null, "operatedByCode": null,
          "from": "JFK", "to": "FLL",
          "departure": "2026-10-24T15:59:00", "arrival": "2026-10-24T19:16:00",
          "durationMinutes": 197 },
        { "flightNumber": "F93442", "carrier": "Frontier Airlines", "carrierCode": "F9",
          "operatedBy": null, "operatedByCode": null,
          "from": "ATL", "to": "MIA",
          "departure": "2026-10-25T08:17:00", "arrival": "2026-10-25T10:15:00",
          "durationMinutes": 118 }
      ] }
  ],
  "itineraryId": "12712-2610241559--32171,-32289-2-14074-2610251015",
  "market": "US", "locale": "en-US", "adults": 1, "childAges": [], "cabinClass": "economy",
  "searchedAt": "2026-09-15T02:53:13.227Z"
}

🧾 Reading the output

The dataset holds itineraries and nothing else. No refusal rows, no notices, no blank lines in your table.

Anything that went wrong goes into a key-value record called PROBLEMS: one entry per search, with the route, the date, a code and a sentence in plain words. It is written on every way out of a run, as an empty list when nothing went wrong, so it is always there. RUN_SUMMARY sits beside it with the counts and your filters written out.

▶️ How to run it

1. Open Skyscanner Flights Scraper and click Try for free. 2. Put your trips in Routes, or paste results links into Skyscanner links. 3. Set Market, Currency and the passengers. 4. Open Filters and order and set Maximum itineraries per search before anything else. 5. Click Start, wait about three minutes, then read the dataset and the PROBLEMS record.

💰 How much does it cost?

$0.052 per 1,000 itineraries, plus $0.005 for each search that finishes. Both rates are the same on every Apify plan, with no volume tiers.

A search that finishes and finds nothing, or finds nothing that passes your filters, still counts as a finished search. A search that is refused, is invalid, could not run, or was cut short before its list finished is not counted, and a cut-short search still delivers the itineraries it had.

Each day of a lastDate range is a search of its own, so a four-day range is four searches.

💡 What people use it for

• Watching one route daily and recording where the cheapest fare moves.
• Comparing a week of departure days at once with lastDate, to find the cheap day to fly.
• Pulling non-stop business fares on a corporate route and feeding them into a travel policy.
• Checking what a competitor's hub pair actually costs before pricing something against it.
• Building a fare history keyed on route and date, with searchedAt as the timestamp.

🚧 What it does not do

• No booking and no seat availability. These are advertised prices and they move. Treat a row as

a quote from the moment in searchedAt.

• No nearby-airport expansion. A link that asks for it gets the named airports only, and says so

in PROBLEMS. Use a city code if you want every airport in a city.

• No whole-month searches. A month link is refused. Use lastDate, up to seven days.
• Two or three legs on a multi-city trip, no more, and lastDate does not apply to them.
• Times carry no time zone. They are local at each airport, exactly as Skyscanner publishes them.
• best is our ranking, not Skyscanner's.
• Prices are for the whole party. Divide it yourself if you want a per-person figure.
• 50 searches in a run. Anything past that is reported in PROBLEMS and left for another run.
• Each search takes about three minutes and cannot be hurried. They do run together.
• No baggage, fare rules, seat maps or loyalty pricing.

🧭 Which travel scraper do you need?

❓ Questions people ask

Do I need a Skyscanner account or an API key? No. Nothing to paste in and nothing to renew.

Why did my run take three minutes for one route? Because the search itself does. Put all your routes in one run: they wait together, so fifty take about as long as one.

Why did I get no rows but a PROBLEMS entry? Either Skyscanner had no priced itineraries (NO_RESULTS) or your filters removed all of them (NO_MATCHES). The reason on the entry says how many it found before filtering.

Can I search a whole month? Not in one go. Use lastDate for a range of up to seven departure days, which is seven searches.

Is the price per person? No. It is the total for everyone on the booking, infants included.

Why is the same flight listed twice? Two rows with the same flight numbers are different itineraries, usually different seller combinations. itineraryId separates them.

🆘 If something breaks

Open the Issues tab on the actor page. Send the run ID and the routes you used, and paste the PROBLEMS record from the run's key-value store. It names the route and the reason for every search that did not deliver.