Each attribute of a Pydantic model has a type. The type can be a built-in Python type or a model itself. Hence it is possible to declare nested JSON “objects” with specific attribute names, types, and validations.

Example

In the following example, we construct a customer model with one of the attributes as product model class. The product model in turn has an attribute of supplier class.

from typing import Tuple
from fastapi import FastAPI
from pydantic import BaseModel
app = FastAPI()
class supplier(BaseModel):
   supplierID:int
   supplierName:str
class product(BaseModel):
   productID:int
   prodname:str
   price:int
   supp:supplier
class customer(BaseModel):
   custID:int
   custname:str
   prod:Tuple

The following POST operation decorator renders the object of the customer model as the server response.

@app.post(''/invoice'')
async def getInvoice(c1:customer):
   return c1

The swagger UI page reveals the presence of three schemas, corresponding to three BaseModel classes.

The Customer schema when expanded to show all the nodes looks like this −

An example response of “/invoice” route should be as follows −

{
   "custID": 1,
   "custname": "Jay",
   "prod": [
      {
         "productID": 1,
         "prodname": "LAPTOP",
         "price": 40000,
         "supp": {
            "supplierID": 1,
            "supplierName": "Dell"
         }
      }
   ]
}