Pydantic V2 & Rust Core Validation

Learning outcomes

• Utilize compiled Rust validation graphs
• Structure FastAPI route schemas efficiently

Mental model

Pydantic V2 shifts data parsing and validation from Python bytecode into a compiled Rust core engine (pydantic-core), delivering 5x-50x faster JSON serialization and schema validation.

Theory

Pydantic V2 compiles Python model definitions into a static C-ABI validation graph during class declaration. When a request arrives, pydantic-core parses JSON directly into C structures, applying type constraints, string regexes, and coercion rules in compiled Rust before instantiating Python objects.

from pydantic import BaseModel, Field, field_validator, EmailStr
from fastapi import FastAPI, status

class UserCreateRequest(BaseModel):
    username: str = Field(min_length=3, max_length=30, pattern=r"^[a-zA-Z0-9_]+$")
    email: EmailStr
    age: int = Field(ge=18, le=120)

    @field_validator("username")
    @classmethod
    def validate_reserved(cls, v: str) -> str:
        if v.lower() in {"admin", "root", "system"}:
            raise ValueError("Username is reserved.")
        return v

app = FastAPI()

@app.post("/users", status_code=status.HTTP_201_CREATED)
async def create_user(user: UserCreateRequest):
    return {"status": "created", "username": user.username}

Alternatives and trade-offs

Pydantic V1 used pure Python reflection, which was significantly slower. Native Python dataclasses lack automatic runtime validation and coercion. Marshmallow requires manual schema definitions. Pydantic V2 combines speed with type hints.

Failure modes and misconceptions

1. Validation vs Verification: Pydantic validates payload syntax and field types; domain business logic (e.g. database uniqueness check) belongs in service layers.
2. model_construct() bypass: Calling User.model_construct() skips Rust validation entirely. Use only for pre-validated internal database objects.

Decision scenario

Use Pydantic V2 models for incoming API request bodies and response DTO schemas. Execute database uniqueness checks in service layers after Pydantic validates structural syntax.

Learning outcomes

• Explain how pydantic-core compiles validation graphs in Rust.
• Implement field validators and field constraints in Pydantic V2.
• Integrate Pydantic models with FastAPI for automated OpenAPI documentation.

Trade-offs

Pydantic V2 offers extreme validation performance, but syntax changes from V1 require migrating validator decorators and config classes.

Primary sources

• Pydantic V2 Architecture & Rust Core Data Validation — Pydantic Inc, verified 2026-07-22
• FastAPI Framework Architecture & Dependency Injection Specification — Tiangolo, verified 2026-07-22