Install: choose the workflow that fits your toolchain

Install the CLI, enable extras, and confirm the pfg entry point is on your PATH.

pip (recommended)

python -m pip install --upgrade pip

pip install pydantic-fixturegen

This installs the CLI script pfg and pulls core dependencies (pydantic, faker, pluggy, typer).
After installing, verify with pfg --help.

Enable optional capabilities per environment:

pip install 'pydantic-fixturegen[orjson]'

pip install 'pydantic-fixturegen[regex]'

pip install 'pydantic-fixturegen[hypothesis]'

pip install 'pydantic-fixturegen[watch]'

pip install 'pydantic-fixturegen[fastapi]'

pip install 'pydantic-fixturegen[polyfactory]'

pip install 'pydantic-fixturegen[openapi]'

pip install 'pydantic-fixturegen[dataset]'

pip install 'pydantic-fixturegen[sqlmodel]'

pip install 'pydantic-fixturegen[beanie]'

pip install 'pydantic-fixturegen[seed]'

pip install 'pydantic-fixturegen[all]'

pip install 'pydantic-fixturegen[all-dev]'

orjson unlocks faster JSON encoding.
regex installs rstr for regex-constrained strings.
hypothesis enables property-based extras in the cookbook.
watch adds filesystem watching via watchfiles.
fastapi pulls in FastAPI + Uvicorn so you can use pfg fastapi smoke and pfg fastapi serve.
polyfactory installs Polyfactory so model discovery can delegate to existing factories and pfg gen polyfactory runs immediately.
sqlmodel installs SQLModel + SQLAlchemy for database seeding.
beanie installs Beanie + Motor for MongoDB seeding.
seed bundles both stacks so pfg gen seed works out of the box.
dataset installs PyArrow so pfg gen dataset --format parquet|arrow works without additional steps.
openapi bundles datamodel-code-generator + PyYAML so you can ingest JSON Schema/OpenAPI documents directly in pfg.
all bundles runtime extras; all-dev adds Ruff, mypy, pytest, and pytest-cov.

poetry add pydantic-fixturegen

poetry run pfg --help

Add extras with poetry add pydantic-fixturegen[watch].
Use poetry run pfg ... inside virtual environments or configure poetry’s virtualenvs.in-project setting for local bins.

# pyproject.toml

[project]

dependencies = ["pydantic-fixturegen"]

hatch run pfg --help

Add extras inside the same dependency list, for example ["pydantic-fixturegen[orjson,watch]"].
Combine with Hatch environments to separate runtime and dev dependencies.

Extra	Provides	Ideal for
`regex`	`rstr` for regex-based string providers	Models with `Field(regex=...)` constraints
`orjson`	Fast JSON encoder	Large JSON/JSONL dumps
`hypothesis`	Property-based strategy exporter (`strategy_for`, `pfg gen strategies`)	Generative testing combos
`watch`	Live regeneration via `watchfiles`	Watch mode in CI or local loops
`numpy`	Deterministic NumPy array providers	Models embedding `numpy.ndarray` fields
`fastapi`	FastAPI + Uvicorn	Smoke tests and mock servers
`polyfactory`	Polyfactory	Delegating to existing `ModelFactory` classes
`sqlmodel`	SQLModel + SQLAlchemy	SQL database seeding via CLI/fixtures
`beanie`	Beanie + Motor	MongoDB seeding
`seed`	SQLModel + SQLAlchemy + Beanie + Motor	Enable all seeding workflows
`dataset`	PyArrow	CSV/Parquet/Arrow dataset emission
`openapi`	`datamodel-code-generator` + PyYAML	JSON Schema / OpenAPI ingestion workflows
`all`	Every runtime extra	One-shot enablement
`all-dev`	Runtime extras + mypy, Ruff, pytest, pytest-cov	Local development stacks

pfg --version

pfg schema config --out /tmp/config.schema.json

pfg --version prints the installed version and resolves the entry point declared under [project.scripts].
pfg schema config confirms Typer loads dynamic subcommands correctly.

You are ready to follow the Quickstart or dig into configuration defaults.