FastAPI Extra Models

Introduction

When building real-world APIs, you'll often need to work with multiple related models rather than just a single data structure. FastAPI provides elegant ways to define, relate, and use multiple Pydantic models to create complex yet maintainable API schemas.

In this tutorial, we'll explore how to work with multiple models in FastAPI, establish relationships between them, and use them effectively in your API endpoints. This is a crucial skill for building robust applications with proper data separation and organization.

Why Use Multiple Models?

Before diving into the code, let's understand why we might need multiple models:

Input vs Output: Often the data structure you receive might differ from what you return
Data Validation: Different endpoints might require different validation rules
Documentation: Multiple models help create clearer, more specific API documentation
Security: You can exclude sensitive fields from response models
Database Integration: Your database models might not match your API models

Basic Multiple Model Example

Let's start with a simple example of using different models for input and output:

python
from fastapi import FastAPI
from pydantic import BaseModel, EmailStr
from typing import List, Optional

app = FastAPI()

class UserIn(BaseModel):
    username: str
    email: EmailStr
    password: str
    full_name: Optional[str] = None

class UserOut(BaseModel):
    username: str
    email: EmailStr
    full_name: Optional[str] = None

@app.post("/users/", response_model=UserOut)
async def create_user(user: UserIn):
    # In a real app, you would save the user to a database here
    # Note that we return the input user, but FastAPI will filter
    # out the password field because we use UserOut as response_model
    return user

In this example:

UserIn includes password and is used for input validation
UserOut excludes password and is used for the response
The response_model=UserOut parameter ensures that FastAPI filters the output to match UserOut

When testing with a request body like:

json
{
  "username": "johndoe",
  "email": "[email protected]",
  "password": "secret123",
  "full_name": "John Doe"
}

The API will return:

json
{
  "username": "johndoe",
  "email": "[email protected]",
  "full_name": "John Doe"
}

Advanced Model Relationships

Real applications often have complex relationships between models. Let's explore some common patterns:

Nested Models

You can define models that contain other models:

python
from fastapi import FastAPI
from pydantic import BaseModel
from typing import List, Optional

app = FastAPI()

class Item(BaseModel):
    name: str
    description: Optional[str] = None
    price: float
    tax: Optional[float] = None

class User(BaseModel):
    username: str
    full_name: Optional[str] = None
    email: str
    
class Order(BaseModel):
    items: List[Item]
    customer: User

@app.post("/orders/")
async def create_order(order: Order):
    return {"order_id": "123", "order": order}

With this structure, your API will expect a request body like:

json
{
  "items": [
    {
      "name": "Laptop",
      "description": "High-performance laptop",
      "price": 999.99,
      "tax": 81.00
    }
  ],
  "customer": {
    "username": "johndoe",
    "full_name": "John Doe",
    "email": "[email protected]"
  }
}

Union Type Models

Sometimes you might need to accept different types of models at the same endpoint. For example, a payment API might handle different payment methods:

python
from fastapi import FastAPI
from pydantic import BaseModel
from typing import Union

app = FastAPI()

class CreditCardPayment(BaseModel):
    card_number: str
    expiry_date: str
    security_code: str
    
class BankTransferPayment(BaseModel):
    account_number: str
    routing_number: str
    
class PaymentResponse(BaseModel):
    payment_id: str
    amount: float
    status: str

@app.post("/payments/", response_model=PaymentResponse)
async def process_payment(payment: Union[CreditCardPayment, BankTransferPayment]):
    # Process payment based on type
    if isinstance(payment, CreditCardPayment):
        # Process credit card
        payment_method = "credit_card"
    else:
        # Process bank transfer
        payment_method = "bank_transfer"
        
    return {
        "payment_id": "pay_123456",
        "amount": 100.50,
        "status": f"Processing {payment_method} payment"
    }

Model Inheritance

Pydantic models support inheritance, which is great for when you have models that share many fields:

python
from fastapi import FastAPI
from pydantic import BaseModel
from typing import Optional, List

app = FastAPI()

class ItemBase(BaseModel):
    name: str
    description: Optional[str] = None
    price: float
    
class ItemCreate(ItemBase):
    tax: Optional[float] = None
    
class ItemInDB(ItemBase):
    id: int
    stock: int
    
@app.post("/items/", response_model=ItemInDB)
async def create_item(item: ItemCreate):
    # Simulate creating an item in database and returning it
    return ItemInDB(
        id=1,
        stock=20,
        name=item.name,
        description=item.description,
        price=item.price
    )

Using Extra Models for Database Operations

When working with databases, it's common to have distinct models for:

Creating records (with validation)
Reading records (for responses)
Updating records (with partial data)
Database models (ORM integration)

Here's an example using SQLAlchemy with FastAPI:

python
from fastapi import FastAPI, Depends, HTTPException
from sqlalchemy.orm import Session
from pydantic import BaseModel
from typing import Optional, List

# Database setup (simplified)
from database import get_db, Base
from sqlalchemy import Column, Integer, String, Float
from sqlalchemy.ext.declarative import declarative_base

app = FastAPI()

# SQLAlchemy model
class ProductDB(Base):
    __tablename__ = "products"
    
    id = Column(Integer, primary_key=True, index=True)
    name = Column(String, index=True)
    description = Column(String)
    price = Column(Float)
    stock = Column(Integer)

# Pydantic models
class ProductBase(BaseModel):
    name: str
    description: Optional[str] = None
    price: float
    
class ProductCreate(ProductBase):
    stock: int
    
class ProductUpdate(BaseModel):
    name: Optional[str] = None
    description: Optional[str] = None
    price: Optional[float] = None
    stock: Optional[int] = None

class Product(ProductBase):
    id: int
    stock: int
    
    class Config:
        orm_mode = True

@app.post("/products/", response_model=Product)
def create_product(product: ProductCreate, db: Session = Depends(get_db)):
    db_product = ProductDB(**product.dict())
    db.add(db_product)
    db.commit()
    db.refresh(db_product)
    return db_product

@app.get("/products/{product_id}", response_model=Product)
def read_product(product_id: int, db: Session = Depends(get_db)):
    db_product = db.query(ProductDB).filter(ProductDB.id == product_id).first()
    if db_product is None:
        raise HTTPException(status_code=404, detail="Product not found")
    return db_product

@app.put("/products/{product_id}", response_model=Product)
def update_product(product_id: int, product: ProductUpdate, db: Session = Depends(get_db)):
    db_product = db.query(ProductDB).filter(ProductDB.id == product_id).first()
    if db_product is None:
        raise HTTPException(status_code=404, detail="Product not found")
        
    update_data = product.dict(exclude_unset=True)
    for key, value in update_data.items():
        setattr(db_product, key, value)
        
    db.commit()
    db.refresh(db_product)
    return db_product

In this example:

ProductBase contains common fields
ProductCreate is used for creating new products
ProductUpdate makes all fields optional for partial updates
Product is the response model with all fields including the ID
ProductDB is the SQLAlchemy ORM model

Using Generic Models

For APIs with common response patterns, you can create generic models to reduce repetition:

python
from fastapi import FastAPI
from pydantic import BaseModel, Generic, TypeVar
from typing import Optional, List, Dict, Any

T = TypeVar('T')

app = FastAPI()

class User(BaseModel):
    id: int
    name: str
    email: str

class Item(BaseModel):
    id: int
    name: str
    price: float

class PaginatedResponse(BaseModel, Generic[T]):
    items: List[T]
    total: int
    page: int
    size: int
    
class GenericResponse(BaseModel, Generic[T]):
    data: T
    status: str
    message: Optional[str] = None
    meta: Optional[Dict[str, Any]] = None

@app.get("/users/", response_model=PaginatedResponse[User])
async def list_users():
    # Simulate database fetch
    users = [
        User(id=1, name="Alice", email="[email protected]"),
        User(id=2, name="Bob", email="[email protected]"),
    ]
    return PaginatedResponse(
        items=users,
        total=2,
        page=1,
        size=10
    )

@app.get("/items/{item_id}", response_model=GenericResponse[Item])
async def get_item(item_id: int):
    # Simulate database fetch
    item = Item(id=item_id, name="Laptop", price=999.99)
    return GenericResponse(
        data=item,
        status="success",
        message="Item retrieved successfully",
        meta={"accessed_at": "2023-07-15T15:30:00Z"}
    )

This approach provides consistent responses across different endpoint types while maintaining type safety.

Summary

Working with multiple models in FastAPI offers several advantages:

Security: Keep sensitive data out of responses
Flexibility: Handle different input and output requirements
Organization: Create logical separations between your data models
Maintainability: Update input validation without affecting output models
Documentation: Generate clear API documentation with specific schemas

By properly using extra models, you can create more robust, secure, and maintainable API designs that handle complex real-world requirements.

Exercises

To reinforce your understanding of extra models in FastAPI:

Create an e-commerce API with models for products, customers, and orders with appropriate relationships between them.
Implement a blog system with models for users, posts, and comments using inheritance to reduce code duplication.
Design an authentication system with separate models for registration, login, and user profiles.
Create a generic pagination response model and apply it to several endpoints.

Additional Resources

If you spot any mistakes on this website, please let me know at [email protected]. I’d greatly appreciate your feedback! :)

Introduction​

Why Use Multiple Models?​

Basic Multiple Model Example​

Advanced Model Relationships​

Nested Models​

Union Type Models​

Model Inheritance​

Using Extra Models for Database Operations​

Using Generic Models​

Summary​

Exercises​

Additional Resources​