Brazil's largest vehicle marketplace as JSON - with both years, and FIPE next to the asking price

Because Brazilian vehicles have an ano de fabricacao (the year the car was built) and an ano do modelo (the model year it is sold as), and they are often not the same. On 2026-09-23 they differed on 41 of 96 sampled listings - 9 of 24 cars, 10 of 24 motorbikes, 14 of 24 private-seller listings and 8 of 24 new cars. The API returns them as year_manufacture and year_model. They are never merged into a single year field, because on nearly half the inventory that would be a wrong value rather than a missing one. The year_min and year_max filters work on the ano do modelo, which is what Webmotors itself filters on.

Yes, and it keeps it apart from the asking price. FIPE is Brazil's reference price index. Every listing carries fipe_price and fipe_code as stored on the ad, plus fipe_percent, which is Webmotors' own figure for how the asking price compares - 88 means the car is listed at 88 per cent of FIPE. The reference_price action returns the live FIPE for that vehicle with its own update date, and alongside it Webmotors' own lowest, average and highest asking price for the same vehicle in that state, with a separate update date. Those two things are named differently on purpose: fipe_price is the index, market_lowest_price / market_average_price / market_highest_price are what people are actually asking. On recent listings the two FIPE figures agree; on older ads the figure stored on the ad can be a few per cent stale, so both are given rather than averaged.

Each one was checked live on 2026-09-23: the filtered total had to be smaller than Webmotors' unfiltered total measured in the same run, and every returned row had to match the filter. Out of 350,975 cars - Toyota 23,519, Toyota Corolla 6,628, 50,000 to 70,000 BRL 44,939, model years 2022 to 2024 95,415, under 20,000 km 83,212, over 100,000 km 66,928, manual 92,146, diesel 33,133, pickups 37,236, white 104,569, two-door 19,789, private sellers 71,373, dealers 179,322, new 40,141, used 310,878, armoured 22,392, air conditioning 276,415, one owner 67,629, Parana 32,198, the city of Sao Paulo 80,837, and a single dealer's stock 41. Filters Webmotors accepts but ignores are deliberately not offered, so a filter you pass always does something.

The five Webmotors itself offers: relevance, price_asc, price_desc, year_desc (newest model year) and mileage_asc (lowest km). Each was verified to actually reorder a full page of results. Webmotors publishes no mileage-descending, year-ascending or newest-listing order, so none is offered rather than accepted and ignored. One warning: Webmotors does not validate the prices sellers type, so price_desc opens with listings in the hundreds of millions of BRL where the seller left out the decimal separator. Combine it with price_max.

Webmotors serves about 10,000 rows per query however large the total is - an unfiltered car search reports 350,975 results but only 417 pages of 24. Every response returns both total_results and page_total, so you can see the ceiling instead of discovering it halfway through a job. To reach the rest of the inventory, narrow the query - by make, by state, by model year - and page each slice. Page size goes up to 100 rows.

Everything the public listing page shows: the trading name, whether they are a dealership, a shop or a private individual, the seller id (which you can feed back into search as dealer_id to pull their whole stock), city, state and state code, neighbourhood, street, postcode and Webmotors' dealer score. The full record adds the CNPJ, the date they joined, their opening hours for all seven days, and the phone numbers Webmotors publishes on the listing, which were present on 8 of 10 records checked. Private individuals have no trading name on Webmotors - that field is empty because the source publishes nothing there, not because it was removed.

Yes - 53,422 motorbikes on 2026-09-23, through the same search action with vehicle_type set to bike, and 87 motorbike makes in makes_models against 138 car makes. Motorbikes have their own specification: engine_cc, engine type, brakes, fuel system, cooling and starter, and they can be filtered by displacement with engine_cc_min and engine_cc_max. They do not have a version, a door count, a power figure or a drivetrain on Webmotors, so those fields are null rather than guessed from the title. Motorbike listings round-tripped 5 of 5 from a search row to the full record.

There is no free-text search on this data - you filter by make, model and version, which is how Webmotors' own search box works. There is no radius search around a point; you filter by state or city instead. A search row carries no fuel type and no publication date - the full listing has both. A private seller has no published name. Motorbikes have no version, doors, power or drivetrain. Engine size is patchy on some segments - it was present on only 11 of 24 hatchbacks and 12 of 24 private-seller rows - and Webmotors' good_deal badge appears on roughly a quarter to a half of rows depending on the category. All of these are returned as null rather than filled in with a guess.

Webmotors API is a ReefAPI endpoint group for brazil's largest car and motorbike marketplace as json, with both brazilian years and the fipe price. It returns live JSON through POST requests under /webmotors/v1.

Yes. ReefAPI starts with 1,000 free credits, no card required. Webmotors calls use the same shared credit balance as every other ReefAPI engine.

No login to Webmotors is needed for the API response. You call ReefAPI with your x-api-key header, and the playground can run live examples before you create a production key.

The page example is captured from a live search call, and production requests fetch live data through ReefAPI rather than a static sample.

Webmotors actions currently cost 1-2 credits per successful call. Failed or blocked calls are free. All APIs draw from one credit pool.

Yes. Connect ReefAPI once through MCP and your assistant can call webmotors actions with the same key, credit pool and JSON envelope used by normal REST requests.