DEV Community

Lasisi Ibrahim Pelumi
Lasisi Ibrahim Pelumi

Posted on

Running FastAPI in Production on a VPS (Step-by-Step)

#webdev #fastapi #productivity #devops

FastAPI makes it super easy to build APIs — but deploying it to real production on a VPS is where many developers get stuck.

Over the past months, I’ve deployed multiple FastAPI services (APIs, webhooks, and automation backends) to VPS servers for real users. In this guide, I’ll walk through a clean, production-ready setup you can reuse for your own projects.

We’ll go from:

main.py on your laptop → Live HTTPS API on a VPS with gunicorn, uvicorn, systemd, and nginx.

What We’ll Build

We’ll deploy a FastAPI app so that:

  • It runs behind gunicorn + uvicorn workers
  • It’s managed by systemd (auto-start on reboot, restart on crash)
  • It’s served via nginx as a reverse proxy
  • It’s protected with HTTPS using Let’s Encrypt
  • It uses environment variables for secrets (no hardcoded tokens)
  • It’s easy to update with a simple git pull + restart

I’ll assume:

  • You’re using Ubuntu 20.04+ on your VPS
  • You have SSH access
  • You have a domain pointing to the VPS IP (optional but recommended)
  1. Create a Minimal FastAPI App

On your local machine (or directly on the server), create a basic FastAPI app:

mkdir fastapiapp
cd fastapiapp
python -m venv venv
source venv/bin/activate   On Windows: venv\Scripts\activate
pip install fastapi uvicorn[standard]
Enter fullscreen mode Exit fullscreen mode

Create app/main.py:

from fastapi import FastAPI

app = FastAPI()


@app.get("/")
async def read_root():
    return {"message": "Hello from FastAPI on VPS!"}
Enter fullscreen mode Exit fullscreen mode

Test locally:

uvicorn app.main:app --reload
Enter fullscreen mode Exit fullscreen mode

If it works locally, we’re ready to move to the VPS.

  1. Connect to Your VPS

From your local terminal:

ssh root@YOUR_SERVER_IP
Enter fullscreen mode Exit fullscreen mode

(Replace root with your username if you’re using a non-root user.)

Update packages:

apt update && apt upgrade -y
Enter fullscreen mode Exit fullscreen mode

Install some essentials:

apt install -y git python3 python3-venv python3-pip nginx
Enter fullscreen mode Exit fullscreen mode
  1. Clone Your FastAPI Project to the VPS

Inside your server:

cd /opt
git clone https://github.com/ibrahimpelumi6142/fastapi-production-guide
cd fastapi-production-guide
python3 -m venv venv
source venv/bin/activate
pip install --upgrade pip
pip install fastapi uvicorn[standard] gunicorn
Enter fullscreen mode Exit fullscreen mode

(If your project is private, use ssh clone or GitHub deploy keys.)

You should now have:

fastapi-production-guide/
├── app/
│   └── main.py
├── nginx/
│   └── fastapi.conf
├── systemd/
│   └── fastapi.service
├── .env.example
├── requirements.txt
└── README.md
Enter fullscreen mode Exit fullscreen mode
  1. Run FastAPI with Gunicorn + Uvicorn Workers

Instead of running plain uvicorn, in production we use gunicorn to manage multiple uvicorn workers.

From /opt/fastapiapp:

source venv/bin/activate
gunicorn -k uvicorn.workers.UvicornWorker app.main:app --bind 0.0.0.0:8000 --workers 3
Enter fullscreen mode Exit fullscreen mode

Your app should be available on port 8000:

curl http://127.0.0.1:8000/
Enter fullscreen mode Exit fullscreen mode

If you see:

{"message":"Hello from FastAPI on VPS!"}
Enter fullscreen mode Exit fullscreen mode

We’re good.

Now we’ll turn this into a service.

  1. Create a systemd Service (So It Runs in the Background)

We don’t want to manually start gunicorn every time.
We’ll use systemd to manage it like a proper service.

Create a service file:

nano /etc/systemd/system/fastapi.service
Enter fullscreen mode Exit fullscreen mode

Paste this (update paths/usernames where needed):

[Unit]
Description=Gunicorn instance to serve FastAPI app
After=network.target

[Service]
User=root
WorkingDirectory=/opt/fastapi-production-guide
Environment="PATH=/opt/fastapi-production-guide/venv/bin"
ExecStart=/opt/fastapi-production-guide/venv/bin/gunicorn -k uvicorn.workers.UvicornWorker app.main:app --bind 0.0.0.0:8000 --workers 3

Restart=always
RestartSec=5

[Install]
WantedBy=multi-user.target
Enter fullscreen mode Exit fullscreen mode

Save and exit.

Reload systemd:

systemctl daemon-reload
systemctl start fastapi
systemctl enable fastapi
Enter fullscreen mode Exit fullscreen mode

Check status:

systemctl status fastapi
Enter fullscreen mode Exit fullscreen mode

If it’s active (running), your backend is now running in the background, even if you log out.

  1. Configure Nginx as a Reverse Proxy

We don’t expose gunicorn directly to the internet. Instead, we let nginx:

  • Handle HTTP(S)
  • Proxy requests to 127.0.0.1:8000
  • Deal with static files (if needed)
  • Improve security & performance

Create a new Nginx config:

nano /etc/nginx/sites-available/fastapi
Enter fullscreen mode Exit fullscreen mode

If you have a domain, use it. If not, you can temporarily use the server IP.

server {
    listen 80;
    server_name YOUR_DOMAIN_NAME_HERE;   or your server IP

    location / {
        proxy_pass         http://127.0.0.1:8000;
        proxy_redirect     off;

        proxy_set_header   Host                 $host;
        proxy_set_header   X-Real-IP            $remote_addr;
        proxy_set_header   X-Forwarded-For      $proxy_add_x_forwarded_for;
        proxy_set_header   X-Forwarded-Proto    $scheme;
    }
}
Enter fullscreen mode Exit fullscreen mode

Enable this site:

ln -s /etc/nginx/sites-available/fastapi /etc/nginx/sites-enabled/
rm /etc/nginx/sites-enabled/default   optional, remove default site
nginx -t
systemctl restart nginx
Enter fullscreen mode Exit fullscreen mode

Now, visit:

  • http://YOUR_DOMAIN_NAME/ or
  • http://YOUR_SERVER_IP/

You should see your FastAPI JSON response.

  1. Add HTTPS with Let’s Encrypt (Certbot)

Never leave a public API on pure HTTP.

Install Certbot:

apt install -y certbot python3-certbot-nginx
Enter fullscreen mode Exit fullscreen mode

Run:

certbot --nginx -d YOUR_DOMAIN_NAME_HERE
Enter fullscreen mode Exit fullscreen mode

Follow the prompts:

  • Provide email
  • Accept terms
  • Choose redirect option → Redirect all traffic to HTTPS

After this, your site is live at:

https://YOUR_DOMAIN_NAME/
Enter fullscreen mode Exit fullscreen mode

Certbot also sets up automatic certificate renewal.

  1. Using Environment Variables for Secrets

Never hardcode secrets like:

  • API keys
  • Database credentials
  • JWT secrets

In your app, use os.getenv or pydantic BaseSettings.

Example: app/config.py

from pydantic import BaseSettings


class Settings(BaseSettings):
    app_name: str = "My FastAPI App"
    secret_key: str
    debug: bool = False

    class Config:
        env_file = ".env"


settings = Settings()
Enter fullscreen mode Exit fullscreen mode

In main.py:

from fastapi import FastAPI
from .config import settings

app = FastAPI(title=settings.app_name)
Enter fullscreen mode Exit fullscreen mode

On the server, create .env in /opt/fastapiapp:

nano /opt/fastapiapp/.env
Enter fullscreen mode Exit fullscreen mode

Example:

SECRET_KEY=supersecretvalue
DEBUG=false
Enter fullscreen mode Exit fullscreen mode

Restart the service:

systemctl restart fastapi
Enter fullscreen mode Exit fullscreen mode

Now your secrets live outside the code.

  1. Updating Your App (Deploying New Versions)

When you change your code on GitHub, deploy updates like this:

ssh root@YOUR_SERVER_IP
cd /opt/fastapiapp
git pull
source venv/bin/activate
pip install -r requirements.txt   if using one
systemctl restart fastapi
Enter fullscreen mode Exit fullscreen mode

That’s it — instant deploy.

If you want smoother deploys, you can:

  • Add a deploy.sh script
  • Use systemctl reload for zero-downtime (with more advanced gunicorn configs)
  • Add logging to file + log rotation
  1. Basic Logging & Monitoring

You can see logs via systemd:

journalctl -u fastapi -f
Enter fullscreen mode Exit fullscreen mode

Or Nginx logs:

tail -f /var/log/nginx/access.log
tail -f /var/log/nginx/error.log
Enter fullscreen mode Exit fullscreen mode

From here, you can integrate:

  • Uvicorn access logs
  • Centralized logging (e.g. Loki, ELK, etc.)
  • Health checks & uptime monitoring (e.g. UptimeRobot or Better Stack)
  1. Common Mistakes to Avoid

Some things I’ve seen (and done myself) that cause pain in production:

❌ Running uvicorn directly in screen/tmux

If the terminal dies → app dies. Use systemd.

❌ Exposing port 8000 to the world

Always hide gunicorn/uvicorn behind Nginx.

❌ No HTTPS

APIs without HTTPS are insecure and sometimes blocked by clients.

❌ Hardcoding tokens in code

Always use .env + BaseSettings or environment variables.

❌ No restart policy

If your app crashes once and doesn’t restart, you’re down. systemd’s Restart=always saves you here.

  1. When to Move Beyond a Single VPS

A single VPS with FastAPI + gunicorn + Nginx is great for:

  • MVPs
  • Side projects
  • Small SaaS
  • Internal tools
  • Bots & automation services

When do you outgrow it?

  • You need auto-scaling
  • You want zero-downtime blue/green deployments
  • You need container orchestration (Kubernetes, ECS, etc.)
  • You have many microservices and need a service mesh

But for 80% of real-world use cases, a well-configured VPS is more than enough, especially when you’re just shipping and proving value.

Conclusion

Deploying FastAPI to production doesn’t have to be scary.

With:

  • gunicorn + uvicorn workers
  • systemd to manage your process
  • nginx as a reverse proxy
  • Let’s Encrypt for HTTPS
  • environment variables for secrets

…you already have a solid, production-ready setup that many teams use in real products.

If you’re building something like a WhatsApp bot backend, job assistant, automation service, or SaaS API, this stack will carry you very far.

About the Author

I’m Lasisi Ibrahim Pelumi, a full-stack developer and automation engineer based in the UK. I build AI-powered products, automation tools, and APIs — including WhatsApp-based assistants and SaaS platforms deployed on VPS and cloud infrastructure.

You can find more of my work on:

  • GitHub: @ibrahimpelumi6142
  • LinkedIn: Lasisi Ibrahim Pelumi

Top comments (0)