The FIPE Table is an excellent research resource for anyone who owns or plans to buy a car. Standardizing the average price of new and used cars greatly facilitates negotiation and, in a way, feeds pricing across the entire automotive market chain.
Because of that, many people dream of having access to this data as an API to embed in their own applications. The problem is that FIPE (yes, that's the foundation's name, not the car price table) doesn't provide this structured data for queries.
There are some initiatives online to turn this data into an API, but they all lack something: up-to-date data, endpoint performance, configuration difficulty, and so on...
Enter my idea => fipe.amb1.io
How the data was acquired
The first step was not to use the existing structure on the FIPE Table website. Python scraping was necessary and the information was downloaded to a MongoDB instance.
Autocomplete search
The way search information is presented felt too rigid — I thought it needed something more organic like autocomplete. Instead of selecting brand, model, and version step by step, you now only need to search for the desired year separately after selecting the vehicle and its version.
Price list
For each vehicle found you can get all values for 2023 starting in January/2023. The site also shows appreciation/depreciation percentages relative to the previous month and the start of the year.
But... where's the API?
The API is available at fipe.amb1.io/api with endpoints to search brands, models, years, and prices — built on Cloudflare Workers for edge performance.