datamodel-code-generator

What it is and what it does

Datamodel-code-generator is a CLI tool and library that converts schema definitions into type-safe Python model code. It reads from OpenAPI 3, AsyncAPI, JSON Schema, Apache Avro, XML Schema, Protocol Buffers, GraphQL, MCP tool schemas, and raw data (JSON/YAML/CSV), then outputs Pydantic v2, Pydantic v2 dataclass, dataclass, TypedDict, or msgspec models. The generated code handles complex schema features like $ref, allOf, oneOf, anyOf, enums, and nested types.

It's designed for developers who need to keep Python models in sync with external schema definitions or who want to avoid hand-writing boilerplate. The tool supports both standalone CLI use and integration as a development dependency in projects. It can also retarget existing Python types (Pydantic, dataclass, TypedDict) to different output formats via the --input-model flag. Output is formatted by default with black and isort, though faster builtin or ruff formatters are available.

Use it for:

• Generate Pydantic models from OpenAPI specs to keep API clients and servers synchronized with schema changes.
• Convert JSON Schema definitions into dataclasses for configuration file parsing with type safety.
• Transform Protocol Buffer definitions into Python models for gRPC service development.
• Generate TypedDict or msgspec models from raw JSON/YAML/CSV data for rapid prototyping.
• Retarget existing Pydantic models to dataclass output for environments that don't use Pydantic.
• Automate model generation in CI/CD pipelines to enforce schema-driven code generation.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Generates Python data models (Pydantic, dataclasses, TypedDict, msgspec) from schema definitions like OpenAPI, JSON Schema, Protocol Buffers, GraphQL, and raw data formats.

Yes. Active maintenance, no known vulnerabilities, low install friction, and permissive MIT license make it a safe choice. The tool solves a real problem—keeping Python models in sync with external schemas—and supports a wide range of input formats. Install as a dev dependency or standalone CLI depending on your workflow.

Install

datamodel-code-generator on PyPI

pip install datamodel-code-generator

uv add datamodel-code-generator

poetry add datamodel-code-generator

Installing datamodel-code-generator

Before you install

Low friction: pure Python wheel with nine runtime dependencies (argcomplete, black, genson, inflect, isort, jinja2, pydantic, pyyaml, tomli). Active maintenance—released 2 days ago with 4002 GitHub stars and commits as recent as 2026-08-14.

License in practice

MIT license (permissive) allows commercial and private use with minimal restrictions; attribution required but no copyleft obligations.

Quickstart

pip install datamodel-code-generator

datamodel-codegen \
  --input schema.json \
  --input-file-type jsonschema \
  --output-model-type pydantic_v2.BaseModel \
  --output model.py

Requires Python 3.10 or later; optional extras (http, graphql, protobuf, ruff) needed for specific schema types or formatters.

Verify before relying

• Performance characteristics for very large or deeply nested schemas
• Completeness of schema feature support across all input formats (e.g., edge cases in $ref resolution)
• Whether the experimental httpx2 backend is production-ready

Package facts

Evidence: datamodel_code_generator-0.72.4-py3-none-any.whl

Tags

More Build Tools packages