Setup

Overview

FastAPI is a Python framework for building APIs. It sits on Starlette, which handles requests, routing and WebSockets, and Pydantic, which validates data. You write ordinary functions with type hints; FastAPI uses those hints to parse and validate input, serialize output, and generate an OpenAPI schema that powers the interactive docs at /docs. The app runs on an ASGI server such as Uvicorn.

Key concepts

ASGI

The async successor to WSGI - the interface between the server and your app. It lets one process serve many concurrent connections, including WebSockets.

Uvicorn

The ASGI server that listens on a port and hands each request to your app. The fastapi dev and fastapi run commands start it for you.

"fastapi[standard]"

Installs FastAPI plus the usual extras: Uvicorn, the fastapi CLI, python-multipart for forms and uploads, email-validator, httpx and Jinja2.

OpenAPI docs

Generated from your code: Swagger UI at /docs, ReDoc at /redoc, and the machine-readable schema at /openapi.json, which clients can generate SDKs from.

Best practices

• Use a virtual environment per project and pin dependency versions.
• Use fastapi dev (auto-reload) while developing; in production run without --reload.
• Start with a single main.py and split into routers once it covers more than one resource.