JSON to Python Converter

Generate Python Models from JSON Samples

Use this JSON to Python converter to create a first draft of Python types from an example API response, configuration file, webhook payload, or test fixture. Choose TypedDict when your code works with dictionaries and you want a type checker to understand their keys. Choose a dataclass when you want named instances with attributes. The generated code is a model of the examples you provide, not a complete description of every value your application may receive.

Paste JSON or open a local file, set a root type name, select a model style, and generate the code. The tool infers nested object models, list item types, nullable values, and keys missing from some sample objects. You can include sample data, then copy the result or download a .py file. Conversion happens in your browser; the JSON is not sent to a conversion service.

How to Convert JSON to Python

1. Paste a valid JSON object or array into the input, or select Open JSON.
2. Enter a valid Python class name, such as Person.
3. Choose TypedDict or Dataclass.
4. For TypedDict output, leave Mark missing fields as optional on if some records omit keys. This control is not used for dataclass output.
5. Choose whether to include a sample dictionary or instance, then select Generate Python.
6. Review the inferred fields and types before copying or downloading the code.

When the input is an array of objects, include a few representative records, especially records that show optional keys, null values, and different nested shapes.

Optional Keys and Nullable Values Are Different

In a TypedDict, NotRequired[T] means a dictionary may omit that key. Optional[T] means the key’s value may be None. A key can be both optional and nullable, which is useful for API fields that sometimes do not appear and may also explicitly contain JSON null.

Sample JSON

[
  { "id": 101, "name": "Ada", "nickname": null },
  { "id": 202, "name": "Grace", "nickname": "Amazing Grace" },
  { "id": 303, "name": "Katherine" }
]

With the root name set to Person, the TypedDict output includes:

class Person(TypedDict):
    id: int
    name: str
    nickname: NotRequired[Optional[str]]

NotRequired reflects that the third record has no nickname key; Optional[str] reflects the explicit null in the first record. Dataclass output handles a missing field with a default of None, but that does not distinguish an omitted key from one that was explicitly null.

TypedDict or Dataclass?

Both styles use Python type annotations. TypedDict instances are ordinary dictionaries at runtime, and their declared key types are not enforced automatically. Dataclasses provide a convenient object shape, but they do not parse JSON dictionaries into instances by themselves.

How JSON Values Map to Python Types

When sample values for a field have incompatible types, the generator may use Any instead of guessing a union. An empty list is typed as list[Any] because it contains no element example. Add representative values or refine those annotations to match the data contract.

How Names and Nested Objects Are Handled

Nested JSON objects become separate named TypedDicts or dataclasses, using class-style names derived from their property names. Dataclass field names are converted to lowercase with underscores where appropriate. TypedDict output preserves the original JSON key; if that key is not a valid Python identifier, the generator uses TypedDict’s functional form so the key can still be represented.

Dataclasses use Python attribute names, so a converted name such as first_name does not automatically map to an external JSON key such as first-name. If you need to serialize or deserialize such data, configure a JSON library or add explicit mapping code.

Python Version Notes

The generated list annotations use built-in generics such as list[str], which require Python 3.9 or newer. When optional TypedDict keys are inferred, the output imports NotRequired from typing, which is available in Python 3.11 and later. For older Python versions, use a compatible typing_extensions import or adapt the generated declaration. Confirm the target Python version for your project before checking the code in.

Generated Types Do Not Validate JSON at Runtime

The generator parses the sample and writes type declarations and optional example data. It does not validate future input, enforce business rules, or turn a JSON dictionary into a dataclass instance. Use a runtime validation or serialization library when data from an API or user must be checked before your program relies on it. If you need a machine-readable JSON contract, start with the JSON Schema Generator and review its inferred requirements.

Frequently Asked Questions

Related JSON Tools

Check your sample with the JSON Validator, make it easier to inspect with the JSON Formatter, or explore nested values in the JSON Viewer. You can also generate other model formats with the JSON to Java Converter and JSON to TypeScript Converter.