PROVISION · DEPLOY · MONITOR

Deploy and manage
on your own servers

Connect a VPS and get push-to-deploy, SSL, workers, observability and more. One flat fee, unlimited apps, no PaaS markup.

Deploy server See features
Hetzner Hetzner
DigitalOcean DigitalOcean
AWS AWS
Custom (SSH)

1Provision a server

One click. We install the whole stack on your VPS.

Ubuntu 24.04 0:12
Python 3.13 0:08
PostgreSQL 17 0:23
Redis 7 0:06
nginx + SSL 0:04
Firewall
Heartbeat

2Add apps

Auto-detects Django, FastAPI, Flask. No Dockerfile, no buildpack, no DSL.

acme/myapp
Add app
Django 5.2 detected
requirements.txt parsed
Let's Encrypt SSL
Already on this server:
api.example.com · blog.example.com

3Push to deploy

Build, migrate, restart. Live metrics + alerts from the first deploy.

$ git push origin main
build 0:18
migrate
healthcheck
live · myapp.com
MONITORING
CPU 12% Mem 24% live

THEN IT RUNS ITSELF

Heartbeat monitoring

Every server pings home every 60s. Alerts the moment one goes quiet.

Background processes

Queue workers, schedulers, consumers - supervised. Restart on crash, survive reboots.

Auto-renewing SSL

Let's Encrypt for every domain. Renews itself - no certificate dance.

Scheduled tasks

Crons + management commands on a schedule, managed from the dashboard.

WHAT IT'D COST

Same app. A fraction of the bill.

One real app: web, a background process, Postgres, and Redis. On a PaaS that's four metered services. On Pyvolt it's one small box you own, plus one flat monthly fee.

Pyvolt
~$16 ∞ apps
Fly.io
~$25 1 app
Render
~$31 1 app
Heroku
~$34 1 app

Each figure runs one app: web + worker + Postgres + Redis (PaaS = each service's smallest production tier). The difference: the PaaS charges per app, so a second app roughly doubles the bill, while Pyvolt manages as many apps as fits on your vps for the same flat fee.

CONFIGURE HOWEVER YOU LIKE

Configure in UI or as code

Set your app up in the dashboard, or commit a pyvolt.toml and manage it from your repo. Same settings both ways, kept in sync: the file wins on deploy, and the fields it owns turn read-only in the dashboard.

Read the pyvolt.toml guide →
Dashboard UI
gunicorn app.wsgi
3.12
manage.py migrate
pyvolt.toml
[app]
serve   = "gunicorn app.wsgi"
release = ["manage.py migrate"]
python  = "3.12"

Zero-downtime deploys

Blue-green releases behind a health gate: new code takes traffic only after it answers healthy, a failed deploy never touches the live version, and rollback is one click.

No Docker required

No Dockerfile, no buildpack, no YAML - Pyvolt runs your Python project directly on the server.

Smart autodiscovery

Point at a repo - we detect your settings, ASGI/WSGI, database config, and static build.

Any process, supervised

A Celery worker, RQ, Dramatiq or just any long-running management command - monitored and restarted on crash.

Scheduled tasks

Manage cron jobs and scheduled commands on a schedule through the dashboard.

Server heartbeats

Every server pings home every 60 seconds. CPU, memory, disk, drift - Slack or Discord alert the moment it's wrong.

Auto-renewing SSL

Let's Encrypt for every custom domain plus wildcard for preview URLs. Renewals run themselves - no certificate dance.

Firewall + security

Firewall rules, intrusion protection, SSH key management - configured during provisioning.

Framework-native deploys

Detects your stack - Django, FastAPI, any Python app with a pyproject or requirements.txt - runs migrations, builds assets, restarts the right WSGI or ASGI worker.

UI + AGENTS

A dashboard for you, an API for your agents

Push to deploy ships your releases. Everything else (provisioning servers, configuring apps, tailing logs) lives in the dashboard, or hand it to Claude, Cursor, or any coding agent through the full CLI and REST API. Scriptable end to end.

Read the CLI docs

Common questions

Everything you need to know before deploying.

How is this different from Render or Heroku?
They host your app on their infrastructure and bill for compute. Pyvolt runs on your own VPS - you pay your cloud provider for the server and us for the management layer. At any scale beyond hobby, you save 3-5x.
Do I need a Dockerfile or special config?
No. No Dockerfile, no Procfile, no buildpack, no YAML. Point us at the repo you already have - we read your dependency manifest (pyproject.toml, requirements.txt, Poetry, or Pipenv), detect your WSGI/ASGI entrypoint, build a virtualenv on your server, and generate the Gunicorn/uvicorn service + Nginx config for you. Config comes from environment variables you set in the dashboard. If we can't infer something (an unusual start command, say), we ask - we don't make you write a spec file.
What about Celery and Channels?
Any queue worker (Celery, RQ, ARQ, Dramatiq, TaskIQ) and its scheduler run as supervised background processes managed through the dashboard, with a cron scheduler for one-off periodic commands. Channels/ASGI is auto-detected - we swap Gunicorn for uvicorn when needed.
What happens if Pyvolt goes down?
Your apps keep running. Pyvolt is the control plane - your code, database, and traffic live on your VPS. Our outage means no dashboard, not a down app.
Can I migrate away?
Yes. Your VPS is yours. Everything is standard Nginx, systemd, PostgreSQL. If you leave, you keep a working, well-configured server.

Ready to ship?

Deploy in five minutes. Free tier, no credit card required.

Start free - no credit card