Models, Validation and Dependencies: An Overview

Two ideas power every FastAPI app. Pydantic models describe data: request bodies, responses and settings are validated and converted from type hints. Dependencies (Depends) describe what an endpoint needs — a database session, the current user, pagination parameters — and FastAPI resolves them per request, caching each dependency within a request. Together they keep path functions short and declarative: the signature says what comes in, what's needed and what goes out.

from typing import Annotated
from fastapi import Depends, FastAPI
from pydantic import BaseModel, Field

app = FastAPI()


class Pagination(BaseModel):
    limit: int = Field(20, ge=1, le=100)
    offset: int = Field(0, ge=0)


async def get_db():
    async with SessionLocal() as session:
        yield session


DB = Annotated[AsyncSession, Depends(get_db)]
Page = Annotated[Pagination, Depends()]


@app.get("/orders")
async def list_orders(db: DB, page: Page, user: CurrentUser) -> list[OrderOut]:
    return await orders.for_user(db, user.id, limit=page.limit, offset=page.offset)

Common pitfalls

• Doing validation manually inside endpoints instead of in models.
• Creating database sessions globally instead of per request.
• Repeating Depends(...) everywhere instead of Annotated aliases.

Key points

• Models describe data
• Dependencies describe requirements
• Annotated aliases keep signatures clean

Further reading: FastAPI: Dependencies