JSON to Python Dataclass

Need the same types in another language? JSON to TypeScript, JSON to Go Struct, JSON to Rust Struct, JSON to C# Class, JSON to Java POJO, JSON to Dart Class. Or validate a document against a JSON Schema.

How to generate Python code from JSON

Paste a JSON object or an array of objects into the input box and press Generate. Optional and nullable fields get a None default, written as X | None or Optional[X]. Pydantic output uses Field(alias=...) for keys that are not valid Python names.

One input, every language

Your JSON stays in the box when you switch the language tabs, and the code is generated again at once. Each language also has its own page, so you can bookmark the one you use.

How types are worked out

• Whole numbers and numbers with decimals are told apart.
• Add more samples with Add sample: a key missing from some samples becomes optional, and null makes a type nullable.
• Arrays of objects are merged into one type, and nested objects become their own types named after the key.
• Keys that are not valid names in the language (such as first-name or 2fa) are renamed and mapped back, never rejected.

The sample JSON in the box above, generated with the default options:

{
  "id": 1,
  "name": "Ada",
  "email": null,
  "tags": ["x"],
  "address": { "city": "Pune", "pin": "411001" },
  "createdAt": "2026-01-15T10:30:00Z"
}

from dataclasses import dataclass


@dataclass
class Address:
    city: str
    pin: str


@dataclass
class Root:
    id: int
    name: str
    tags: list[str]
    address: Address
    created_at: str
    email: str | None = None

• A single sample cannot show which fields are optional. Paste two or three different responses for a realistic result.
• A value that is always null gives no type information, so it is treated as a nullable string. Change it by hand if it is something else.
• Numbers are typed from the samples: 1 and 1.0 look the same in JSON, so use a sample with decimals where you need a float.