json-schema-to-pydantic-rs 0.1.0

Fast JSON Schema to Pydantic v2 model generation, powered by Rust

json-schema-to-pydantic-rs

Fast JSON Schema to Pydantic v2 model generation, powered by Rust.

A high-performance drop-in replacement for json-schema-to-pydantic using a Rust core via PyO3.

Performance

5-10x faster than the pure-Python original:

Schema	Original	Rust	Speedup
Simple object (4 fields)	993 us	172 us	5.8x
Nested 3 levels	2.49 ms	306 us	8.1x
With $ref definitions	1.59 ms	164 us	9.7x
oneOf / anyOf / allOf combiners	2.88 ms	656 us	4.4x
Wide object (50 fields)	2.37 ms	354 us	6.7x
Enums and arrays	452 us	69 us	6.6x

Installation

pip install json-schema-to-pydantic-rs

Requires Python 3.10+ and Pydantic v2.10.4+.

Quick start

from json_schema_to_pydantic_rs import create_model

User = create_model({
    "type": "object",
    "properties": {
        "name": {"type": "string", "minLength": 1},
        "age": {"type": "integer", "minimum": 0},
        "email": {"type": "string", "format": "email"},
    },
    "required": ["name", "age"],
})

user = User(name="Alice", age=30, email="[e-mail]")
print(user.model_dump_json())
# {"name":"Alice","age":30,"email":"[e-mail]"}

# Round-trip
restored = User.model_validate_json('{"name":"Alice","age":30}')

API

create_model(schema, **options)

Converts a JSON Schema into a Pydantic BaseModel subclass. Accepts a dict or a JSON str. Supports the full JSON Schema spec: types, formats (uuid, date-time, email, ...), $ref/$defs, allOf/anyOf/oneOf, enum, const, additionalProperties, constraints, and nullable types.

from json_schema_to_pydantic_rs import create_model

# From a dict
Order = create_model({
    "type": "object",
    "properties": {
        "customer": {"$ref": "#/$defs/Customer"},
        "items": {"type": "array", "items": {"$ref": "#/$defs/Product"}, "minItems": 1},
    },
    "required": ["customer", "items"],
    "$defs": {
        "Customer": {
            "type": "object",
            "properties": {
                "name": {"type": "string", "minLength": 1},
                "email": {"type": "string", "format": "email"},
            },
            "required": ["name"],
        },
        "Product": {
            "type": "object",
            "properties": {
                "name": {"type": "string"},
                "price": {"type": "number", "minimum": 0},
            },
            "required": ["name", "price"],
        },
    },
})

order = Order(
    customer={"name": "Alice", "email": "[e-mail]"},
    items=[{"name": "Widget", "price": 9.99}],
)
print(order.customer.name)   # "Alice"
print(order.items[0].price)  # 9.99

# Or from a JSON string
Order = create_model('{"type": "object", "properties": {"name": {"type": "string"}}}')

Options:

create_model(
    schema,
    base_model_type=MyBase,             # Base class for generated models
    predefined_models={                  # Plug your own classes for $ref paths
        "#/$defs/Address": Address,
    },
    allow_undefined_array_items=True,    # Arrays without "items" -> List[Any]
    allow_undefined_type=True,           # Schemas without "type" -> Any
    populate_by_name=True,               # Access fields by both name and alias
)

PydanticModelBuilder

For reusing config across multiple schemas:

from pydantic import BaseModel
from json_schema_to_pydantic_rs import PydanticModelBuilder

class Address(BaseModel):
    street: str
    city: str

builder = PydanticModelBuilder(
    base_model_type=BaseModel,                         # custom base class
    predefined_models={"#/$defs/Address": Address},    # reuse existing models
)
Model = builder.create_pydantic_model(schema)

Non-standard schema properties are preserved via json_schema_extra:

Model = create_model({
    "type": "object",
    "properties": {"bio": {"type": "string", "ui_widget": "textarea"}},
})
Model.model_fields["bio"].json_schema_extra  # {"ui_widget": "textarea"}

Development

uv sync --dev
uv run maturin develop --release
uv run pytest
uv run python bench.py

Compatibility

• Python 3.10 - 3.14
• Pydantic v2.10.4+
• Rust edition 2024

License

MIT — see LICENSE.

Based on json-schema-to-pydantic by Richard Gyiko.