This is the Pydantic v2 textbook. Pairs with FastAPI and the SQLAlchemy textbooks .
Pydantic v2 mental model: validation (input → model), serialization (model → output), the Rust core, and what changed from v1.
Field declaration, types (str, int, datetime, UUID, Decimal, EmailStr, HttpUrl), constraints, defaults, aliases.
All Pydantic v2 validation hooks: field_validator, model_validator, mode=before/after, custom validators via Annotated.
Output side: model_dump, model_dump_json, field/model serializers, exclude/include, mode=python/json, by_alias.
Composing models: nesting, generics with TypeVars, discriminated unions, recursive models.
Custom types via Annotated, RootModel for primitive-typed root, TypeAdapter for arbitrary types, dataclass integration.
When Pydantic coerces vs raises: strict mode, per-field strict, lax/strict tradeoffs at API boundaries.
JSON Schema from Pydantic: model_json_schema, OpenAPI integration, custom schemas, examples.
Pydantic-settings: typed env config, .env, secrets directory, sources, CLI integration.
Pydantic v2 perf tips, hot path optimization, FastAPI patterns, and when to consider msgspec / attrs.