What are the best Python asynchronous web frameworks?

Options
Considered

User
Recs.

May 9, 2023

Last
Updated

FastAPI vs Aiohttp vs Sanic Framework

vs Starlette vs Quart vs Tornado vs Twisted

Specs

Platforms

Cross-platform

License

MIT

Type

standard and feature rich micro-framework

Initial Release

2019

Written in

Python

Default ORM

SqlAlchemy

WEB SERVER

Starlette, based on ASGI (uvicorn)

WEBSITE

https://fastapi.tiangolo.com/

TYPE DECLARATION

you can use python type hints (optional)

Platforms

License

Type

Initial Release

Written in

Default ORM

WEB SERVER

WEBSITE

TYPE DECLARATION

Platforms

License

Type

Initial Release

Written in

Default ORM

WEB SERVER

WEBSITE

TYPE DECLARATION

Platforms

License

Type

Initial Release

Written in

Default ORM

WEB SERVER

WEBSITE

TYPE DECLARATION

Platforms

License

Type

Initial Release

Written in

Default ORM

WEB SERVER

WEBSITE

TYPE DECLARATION

Platforms

License

Type

Initial Release

Written in

Default ORM

WEB SERVER

WEBSITE

TYPE DECLARATION

Platforms

License

Type

Initial Release

Written in

Default ORM

WEB SERVER

WEBSITE

TYPE DECLARATION

Pros + Cons

Pro
Async IO / optional

It's based on Async IO, which gives it high concurrency. But you can use non-async libraries and it runs them appropriately.

Comment

Pro
Automatic docs

It generates interactive API documentation automatically from your code.

Comment

Pro
Data validation

Pydantic validates the data using the types you declared. Even in deeply nested JSON requests.

Comment

Pro
Standards

It is based on standards: OpenAPI, JSON Schema and OAuth 2.0.

Comment

Pro
Editor completion

It is based on Python type declarations, so, editors and tools can give great support. Including type checks and autocompletion everywhere.

Comment

Pro
Dependency injection

It has a simple but powerful dependency injection system, it can be used to handle authentication, per-user rate limiting, authorization controls (e.g. with roles), etc.

Comment

Pro
Database independent

It's independent of database or ORM, but compatible with all of them. Including relational databases and NoSQL.

Comment

Pro
High-performance

It's based on Starlette, so, it's one of the fastest Python frameworks.

Comment

Pro
World class documentation

It has some of the best documentation of any framework.

Comment

Pro
Background tasks

Included support for background tasks, thanks to being based on Starlette.

Comment

Pro
One of the fastest growing communities

[add description]

Comment

Pro
Steady growth

[add description]

Comment

Con
Smaller community

Since FastAPI is relatively new, its community is smaller than Django Rest Framework.

Comment

Pro
API based not on HTML-base like Flask

[add description]

Comment

Pro
WebSockets

Because it's an async framework, it can handle async-native protocols like WebSockets.

Comment

Pro
Deploys in minutes instead of hours

[add description]

Comment

Pro
Fast development

Makes it awesome for machine learning projects.

Comment

Pro
OAuth 2.0

It has integrated support for OAuth 2.0. Including declaring required scopes per endpoint. So, you can easily integrate it with external OAuth 2.0 providers or build your own with it.

Comment

Pro
Fast is really fast

It's easy to develop API based applications in Python on deadlines for Android and IOS Development.

Comment

Pro
Supports GraphQL

Python's graphene library is included as an optional dependency meaning that GraphQL API's are supported out of the box, with no additional tweaking needed.

Comment