FastAPI

FastAPI on Vercel

FastAPI is a modern, high-performance, web framework for building APIs with Python based on standard Python type hints. You can deploy a FastAPI app to Vercel with zero configuration.

Get started with FastAPI on Vercel

You can quickly deploy a FastAPI application to Vercel by creating a FastAPI app or using an existing one:

Deploy FastAPI to Vercel

Deploy Live Example

Get started with Vercel CLI

Get started by initializing a new FastAPI project using Vercel CLI init command:

terminal

vc init fastapi

This will clone the FastAPI example repository in a directory called fastapi.

Exporting the FastAPI application

To run a FastAPI application on Vercel, define an app instance that initializes FastAPI at any of the following entrypoints:

app.py
index.py
server.py
src/app.py
src/index.py
src/server.py
app/app.py
app/index.py
app/server.py

For example:

src/index.py

from fastapi import FastAPI
 
app = FastAPI()
 
@app.get("/")
def read_root():
    return {"Python": "on Vercel"}

Serving static assets

To serve static assets:

Place them in the public/** directory. They will be served as a part of our Edge Network using default headers unless otherwise specified in vercel.json.
Use app.mount() to mount the public directory.

app.py

from fastapi import FastAPI
from fastapi.staticfiles import StaticFiles
from fastapi.responses import FileResponse, HTMLResponse
 
app = FastAPI()
 
app.mount("/public", StaticFiles(directory="public"), name="public")
 
@app.get("/favicon.ico", include_in_schema=False)
async def favicon():
    return FileResponse("public/vercel.svg")

When you deploy a FastAPI app to Vercel, the application becomes a single Vercel Function and uses Fluid compute by default. This means your FastAPI app will automatically scale up and down based on traffic.

Limitations

All Vercel Functions limitations apply to FastAPI applications, including:

Application size: The FastAPI application becomes a single bundle, which must fit within the 250MB limit of Vercel Functions. Our bundling process removes __pycache__ and .pyc files from the deployment's bundle to reduce size, but does not perform application bundling.