Why choose FastAPI

Tram Ho

3 years ago

Introduction

Recently, due to the growing need for microservice development, I mainly code Python – Backend, so I was allowed to choose a framework to develop new projects for the company, after considering between 3 popular frameworks. At using Python as Django, Flask and FastAPI, I decided to choose FastAPI to develop the project, here is the reason for this.

Main Reason

FastAPI is a fairly new micro framework, just released in 2018. The Github framework of this framework is currently reaching 25.4k stars as of January 2021, but due to the application of many new technologies, FastAPI has a few points. Strong, but I feel quite suitable for use in project development:

High performance Due to being based on two fairly strong libs at the moment of python, Pydantic and Starlette, FastAPI has the highest performance out of all current Python frameworks, you can refer to the performance comparison between frameworks. at https://www.techempower.com/
Development Speed With built-in Swagger interface – OpenAPI comes with a fairly simple code, so programmers can release the function very quickly and still have full documentation, this is the most important advantage of FastAPI compared to with other frameworks. Below is a code that prints the healthcheck text

import uvicorn
from fastapi import FastAPI
from pydantic import BaseModel


class ResponseSchemaBase(BaseModel):
    code: str = '00'
    message: str = ''

app = FastAPI()

@app.post("/health-check", response_model=ResponseSchemaBase)
async def health_check():
    return {"message": "Health check success"}

if __name__ == "__main__":
    uvicorn.run(app, host="0.0.0.0", port=8000)

import uvicorn

from fastapi import FastAPI

from pydantic import BaseModel

class ResponseSchemaBase(BaseModel):

code: str = '00'

message: str = ''

app = FastAPI()

@app.post("/health-check", response_model=ResponseSchemaBase)

async def health_check():

return {"message": "Health check success"}

if __name__ == "__main__":

uvicorn.run(app, host="0.0.0.0", port=8000)

Asynchronous Currently asynchronous has been supported since Django 3.x but since release, FastAPI has supported Async developer by default, so FastAPI can only be used with python3.6 and above.

Compare Django, Flask and FastAPI

Below is a comparison table between the 3 Frameworks Django, Flask and FastAPI based on my own experience in experiencing all 3 frameworks.

	Django	Flask	FastAPI
Community	Django’s community is currently quite large and long-standing, so the number of lib support for Django is arguably the most at the moment.	The current Flask community is also large but not as large as Django, so the number of libs is also limited than Django	FastAPI’s community is relatively new, so there are few libs that support it
Python version	All version: 2x, 3x	All version: 2x, 3x	Only python3.x
Performance	Due to support for many features, Django’s perfomance is not high	Performance is quite high	The highest performing framework in current python frameworks
ORM	Django ORM	Flask-sqlachemy	Sqlachemy
Async	Support from Django 3.x	Not yet supported	Completely compatible
Builtin Admin UI	Have	Are not	Are not
Project Frame	Yes, just run python manage startapp yourapp	Are not	Are not
Python shell	Have	Have	No – this is a limitation of a microframework compared to real frameworks
OpenAPI document	The most common is to use Django-rest-framework, with Swagger use django-rest-swagger with python2.x, use drf-yasg with python3.x	Use Flask-restplus	Swagger and Redocs are included by default

Share the news now