Skip to content
FlightPowers
Get key

Guide

How to get Google Flights prices with Python

Published September 5, 2026

Short answer: FlightPowers' Google Flights API answers this with one POST to google-flights-live-api.p.rapidapi.com/api/google_flights/oneway/v1 using Python's requests library, about twelve lines total, no Playwright or browser pool. Every fare in the JSON list already carries price_insights_low, price_insights_high, and a low, typical, or high verdict, so you get price context on the first call.

One POST to a JSON endpoint. Twelve lines, no browser.

There are two ways to do this and people usually try the hard one first. The hard one is driving a real browser: Playwright, a headless Chrome, ARIA selectors, and a result list that streams in after the page says it's loaded. It works. We wrote the whole scraper so you can see what it costs to keep alive.

The other way is one POST and a JSON body.

How do you make the request?

import os, requests

r = requests.post(
    "https://google-flights-live-api.p.rapidapi.com/api/google_flights/oneway/v1",
    headers={
        "Content-Type": "application/json",
        "x-rapidapi-host": "google-flights-live-api.p.rapidapi.com",
        "x-rapidapi-key": os.environ["RAPIDAPI_KEY"],
    },
    json={
        "from_airport": "LHR",
        "to_airport": "BCN",
        "departure_date": "2026-11-18",
        "limit": 2,
    },
)
r.raise_for_status()

for f in r.json():
    print(f["price"], f["airline"], f["duration"], f["price_range_in_relation_to_other_periods"])

Airport codes, a date, an optional cap on rows. No session to warm up, no cookie jar, and nothing to encode by hand. If you have ever tried to build a tfs= parameter yourself, here is why you shouldn't.

You need a key to run it. Grab one from the RapidAPI listing, free plan included, and it works on the next request.

What does the response contain?

The response is a plain list of fares. This is a real answer to the exact request above, run on 2026-09-04. Fares move by the minute, so yours will not match.

[
  {
    "price_range_in_relation_to_other_periods": "typical",
    "price_insights_low": 50,
    "price_insights_high": 110,
    "from_airport": "London (LHR)",
    "to_airport": "Barcelona (BCN)",
    "departure_date": "2026-11-18",
    "price": "$63",
    "price_as_number": 63,
    "duration": "2 hr 15 min",
    "duration_seconds": 8100,
    "airline": "Vueling | Iberia, British Airways",
    "stops": 0,
    "stops_info": [],
    "departure_description": "8:50 PM on Wed, Nov 18",
    "arrival_description": "12:05 AM on Thu, Nov 19",
    "buy_link": "https://www.google.com/travel/flights?tfs=GjwSCjIwMjYtMTEtMTgiIAoDTEhSEgoyMDI2LTExLTE4GgNCQ04qAlZZMgQ3NjQzagUSA0xIUnIFEgNCQ05CAQFIAZgBAg&curr=usd"
  }
]

Two things in there are easy to skim past.

price_as_number is the fare as an integer, which is the field you sort and compare on. price is the formatted string for display. Use the wrong one and your sort is alphabetical, which puts $1,004 under $99.

And the first three fields are Google's own price context: the usual range for this route and date, plus a low, typical or high verdict on the fare in front of you. That is the part you would otherwise build by polling a route for six weeks. There is a longer write-up in is this flight price good.

What are the limits?

The band is not always there. On thin routes and dates far out, Google has no range, so price_insights_low and price_insights_high come back null. The verdict can arrive as an empty string on that path rather than null, so compare it against the three strings you want instead of testing for absence.

Two settings remove the band silently. Sending sort_type: "Price" and sending max_price both select Google's price-sorted page, which carries no insights block. You get a normal 200 with normal fares and a null band. Sort and filter in Python instead.

An empty list is ambiguous unless you read the headers. [] can mean no flights or a failed fetch, and they need different handling. Check x-search-status before you store or alert on anything. The taxonomy is in handling empty flight search results.

It cannot book. This returns shoppable fares and a buy_link that reopens the exact itinerary on Google. Issuing a ticket needs a GDS or a booking platform, and that is a different product.

Twelve lines instead of a browser pool

Live Google Flights fares as JSON, with Google's own price band on every result. Free tier on RapidAPI, no card to try it.

Get a key on RapidAPI →See pricing

Free tier: 10 requests/month. No card to try.