PUBLIC API / V1
Car data.
Zero friction.
Brands, models, years and logos.
A small, useful car database. One JSON request away.
No API keyNo sign-upCORS enabled
TAKE IT FOR A SPIN
GET → application/jsonconst response = await fetch(
'https://carsjson.pages.dev/v1/brands/abarth/models.json'
);
const models = await response.json();[
{
"id": "500",
"name": "500",
"cyrillic_name": "500",
"year_from": 2008,
"year_to": 2026,
"brand_id": "ABARTH",
"nameplates": [
"500",
"500e"
]
}
]
01 / ENDPOINTS
Five routes.
Everything you need.
Use a single brand, fetch a logo, or take the whole dataset with you.
BASE URL
https://carsjson.pages.dev/v1/brands.jsonAll brands, names, years and logo paths. Returns an array.
GET/v1/brands/{brand}/models.jsonEvery model for one brand. Use a lowercase brand ID, like abarth or alfa_romeo.
GET/v1/all.jsonThe complete dataset: an object containing brands and models arrays.
GET/v1/meta.jsonSchema version, data fingerprint, build timestamp and record counts.
GET/logos/{brand}.webpA brand logo, served as a WebP image.
All routes are public GET requests. Successful responses use HTTP 200. Missing paths return HTTP 404; the error body is HTML, so check response.ok before parsing JSON.
02 / DATA FORMAT
Just the
useful fields.
Predictable objects, readable names and the details you need to build a car selector.
Keep IDs as strings. A model ID is unique within its brand; identify a model with the pair brand_id + id.
| FIELD | TYPE / DESCRIPTION |
|---|---|
| id | stringModel identifier, such as 500 or 124_SPIDER. |
| brand_id | stringParent brand identifier, such as ABARTH. |
| name | stringDisplay name from the source. |
| cyrillic_name | stringCyrillic name or localized spelling. |
| year_from | number | nullFirst year listed in the source. |
| year_to | number | nullLast year listed in the source. |
| nameplates | string[]Variant names, for example ["500", "500e"]. An empty array means none are listed. |
Unknown years are null. Year ranges are copied from the source; a recent year does not necessarily mean a model is still in production.
03 / QUICK START
From JSON to your app.
- [01]
Choose a brand
Fetch
/v1/brands.jsonto populate a dropdown. Usenamefor the label andidfor the value. - [02]
Load its models
Lowercase the selected brand ID, then request
/v1/brands/{brand}/models.json. Each model includes its brand ID. - [03]
Make it yours
Filter by name or year in your app. For offline use or cross-brand search, download
/v1/all.jsonand keep a local copy.
04 / GOOD TO KNOW
A few
practical notes.
Simple to request.
Easy to understand.
Do I need an API key?
No. Every endpoint is public and free to access. Make a normal GET request without authorization headers. CORS is enabled, so you can fetch data directly from a browser.
How do I know when the data changes?
Read /v1/meta.json and compare data_version, a SHA-256 fingerprint of the public data and logos. Responses are cached for five minutes. generated_at records the build time, not when the source was collected. There is no fixed update schedule.
Can I search, filter or paginate requests?
The API serves complete JSON files. Query parameters do not filter the response. Fetch a brand's models or the full snapshot and apply search, year filters or pagination in your own application.
How do I use a brand logo?
Each brand has a logo path such as /logos/abarth.webp. Resolve it against the API origin: new URL(brand.logo, 'https://carsjson.pages.dev').href. The result can be used as an image's src.
Where does the data come from?
The catalog is compiled and normalized from public automotive directories into a consistent, developer-friendly JSON format. CarsJSON is an independent open project created by Egor Volchenko and is not affiliated with any listed car brands.
Less setup. More building.
Your next car selector starts with a GET request.