JSON to Python Converter

About this tool

Paste a JSON sample and get Python type definitions you can use immediately. Emits `@dataclass` definitions with typing hints. Optional fields default to `None`.

When you would use it

• Modelling a payload without hand-writing `__init__`
• Bootstrapping models from a real JSON response instead of writing them by hand
• Keeping types in step with an API that has changed

Worked example

{
  "id": 4815,
  "name": "Ada Lovelace",
  "active": true,
  "score": 91.5,
  "roles": ["admin", "editor"],
  "address": { "city": "London", "postcode": "NW1" }
}

from dataclasses import dataclass, field
from typing import Any, List, Optional
@dataclass
class Address:
    city: str
    postcode: str
@dataclass
class Root:
    id: int
    name: str
    active: bool
    score: float
    roles: List[str] = field(default_factory=list)
    address: 'Address'

Things worth knowing

• Every object in an array is merged when inferring the shape, so a field missing from one item is marked optional rather than assumed to always be present.
• Identical nested shapes are de-duplicated into a single type instead of generating near-duplicates.
• Array element types are named from the singular of the field name, so posts produces a Post type.
• A sample with more records produces better types, because optional fields only reveal themselves across several examples.