Flight Fares

What this dataset is

This dataset tracks daily airfares for domestic Brazilian flights, collected systematically following the methodology used by IBGE to calculate the air transport component of the IPCA (Índice Nacional de Preços ao Consumidor Amplo), Brazil’s official inflation index.

The scrapers capture fares across airlines (via Google Flights and direct sources such as GOL and LATAM), looking ~60 days ahead for Saturday departures with Sunday return after 8 days. This matches the exact design of the inflation basket for the airline sub-index, ensuring the data reflects what households face when purchasing leisure air travel in Brazil.

Why it matters

Inflation modeling: Airfares are a volatile component of the IPCA. Analysts and traders use this data to anticipate upside/downside surprises in official inflation releases, which can move rates, swaps, and sovereign curves.
Macro strategy: Flight prices often respond to fuel costs, FX rates, and demand shocks (holidays, strikes, new regulation). Monitoring them gives early signals on Brazilian inflation dynamics and consumer spending.
Market positioning: Hedge funds, asset managers, and local banks integrate airfare data into inflation models (breakevens, NTN-B pricing, Selic expectations). For anyone trading DI futures or Brazilian government bonds, having a forward-looking view of airfares is a competitive edge.

Key features

Granularity: Daily fare quotes across routes, with flight-level detail (airline, departure/arrival airports, time, segments, fare type, carry-on policy).

Consistency: Collected systematically using the IBGE/IPCA methodology.

Longevity: Historical series building since Septemeber/2025; ongoing daily collection ensures a growing time series for backtesting.

Long-term considerations

Methodology stability: IBGE’s collection method is stable but can change—monitoring adjustments is key to keeping this dataset aligned with official IPCA.

Airline practices: Airlines constantly update fare structures (fare classes, baggage rules). Normalization is critical to keep comparability.

Volatility caveat: Airfares are notoriously noisy. Professionals typically aggregate (weekly/monthly averages) or apply smoothing before feeding into models.

We are contantly monitoring these aspects and adapting the data to reflect these caveats.

⚠️ Note on Possible Duplicate Flights

Our flight fare dataset is built from multiple scrapers that may overlap in the flights they capture. For example, we currently run:

Direct airline scrapers — e.g., a scraper that queries the GOL website.

Aggregator scrapers — e.g., a scraper that queries Google Flights.

Because both sources can report the same underlying flight, it is possible to have duplicate entries for the same flight.

For instance, if our GOL scraper finds a flight on 01 Jan 2026 from SDU → GRU departing at 14:00, our Google Flights scraper may also return that exact same flight. These will appear as two distinct records because they come from different scrapers.

This duplication is intentional — we preserve the raw capture from each source to maintain transparency and allow users to verify and compare fares between different providers. If your use case requires unique flights only, you should plan to deduplicate based on the available flight attributes (e.g., company, departure_airport, arrival_airport, departure, departure_time, fare_type, carry_on, and segments).

Here’s a clear and professional way to explain this in your documentation. I’ll keep it in the same tone and depth as the previous note so it fits naturally in your API docs and “Datasets Available” page.

🛫 Note on Round-Trip Pricing & Flight Pairing (IPCA Methodology)

This dataset follows the official IBGE IPCA methodology for measuring airfares. The IPCA assumes ticket prices come from a round-trip purchase, not two one-way tickets.

To reproduce the IPCA approach:

Select two flights from the same airline:

Outbound: any flight departing on a Saturday (e.g., SDU → GRU).

Return: any flight from the same airline on the following Sunday, 8 days after the outbound with the reversed route (e.g., GRU → SDU).

Interchangeable pairing:
All flights operated by the same airline on the Saturday can be paired with any of its return flights on the following Sunday (8 days later). They are considered interchangeable for price measurement purposes.

⚠️ Important: Prices in the dataset are those of round-trip tickets. Buying the same legs separately as one-way flights usually costs more. This reflects IBGE’s official inflation measurement approach.

🎫 Note on `fare_type` for Google Flights Data

When flights are collected through the Google Flights scraper, the fare type (e.g., Economy, Business) is not provided by Google.
For these records, the fare_type field will always be set to UNKNOWN.

⚠️ Important: Even though the specific fare type is unknown, the price already reflects the fare class returned by Google at the time of scraping. In other words, the price is valid and comparable but the exact class (Economy/Business/Promo/etc.) cannot be identified.

What this dataset is

Why it matters

Key features

Long-term considerations

⚠️ Note on Possible Duplicate Flights

🛫 Note on Round-Trip Pricing & Flight Pairing (IPCA Methodology)

🎫 Note on `fare_type` for Google Flights Data

Request

Responses

Flight Fares

What this dataset is#

Why it matters#

Key features#

Long-term considerations#

⚠️ Note on Possible Duplicate Flights#

🛫 Note on Round-Trip Pricing & Flight Pairing (IPCA Methodology)#

🎫 Note on fare_type for Google Flights Data#

Request

Responses

What this dataset is

Why it matters

Key features

Long-term considerations

⚠️ Note on Possible Duplicate Flights

🛫 Note on Round-Trip Pricing & Flight Pairing (IPCA Methodology)

🎫 Note on `fare_type` for Google Flights Data