qioi@ca-central:~$ qioi deploy openwebui --plan m --region ca-central
[ok] payment confirmed · order qio-8f2a41
[ok] 4 vCPU / 8 GB provisioned · ca-central
[ok] private network attached · firewall applied
[ok] openwebui:v0.6-stable installed · signature verified
[ok] workspace.qioi.app → 203.0.113.24 · ssl issued
[ok] health checks passing · backups scheduled daily 03:00 utc
// service ready in 3m 41s — no servers, containers, domains, ssl,
// monitoring, backups, or updates configured by hand
Deploy private AI services
in one click.
QIOI provisions and operates the infrastructure underneath tested AI applications — so a working service is what you get, not an empty server.
state 01 provisioning_server
You need the service, not another empty server.
A traditional VPS hands you a blank machine and a bill. QIOI hands you a running application — monitored, backed up, and kept current. The difference, as a diff:
Small VPS
2 vCPU · 4 GB RAM · 80 GB NVMe
from $12/mo
AI-ready VPS
4–8 vCPU · 8–32 GB RAM · NVMe
from $29/mo
GPU instance
Dedicated GPU · regional availability
contact for capacity
Managed server
Any plan + QIOI operations layer
from $49/mo
Backup storage
Encrypted, off-instance
from $5/mo
Private networking
Isolated networks between your instances
from $9/mo
state 02 configuring_network
Your providers, your regions — attached, never imposed.
QIOI sits above the infrastructure layer through provider adapters. We run on proven external providers today and on SpaceOne capacity when it is available. No single point of dependency.
hetzner[attached]primary capacity
ovhcloud[attached]eu capacity
aws[attached]on demand
vultr[attached]on demand
digitalocean[attached]on demand
spaceone[standby]ecosystem capacity — coming online
state 03 installing_application
The catalog is a registry, not a marketplace.
A small number of well-tested templates beats thousands of unverified apps. Every entry is version-pinned, with defined backups, health checks, and a rollback path.
| package | category | min ram | backups | from | status | |
|---|---|---|---|---|---|---|
| openwebui | ai workspace | 4 gb | daily | $39/mo | managed | show → |
| n8n | workflow automation | 4 gb | daily | $49/mo | managed | show → |
| deepfryr | knowledge & rag | 8 gb | daily | $89/mo | managed | show → |
| openclaw | ai workspace | 8 gb | daily | $69/mo | managed | show → |
| hermes | ai workspace | 8 gb | daily | $69/mo | managed | show → |
| litellm-gateway | model infrastructure | 4 gb | daily | $45/mo | managed | show → |
| flowise | developer environment | 4 gb | daily | $45/mo | managed | show → |
state 04 configuring_domain · issuing_ssl
Your instance. Your data. Your choice of models.
isolation
dedicated or isolated instances — no shared runtime with other customers
model keys
your provider api keys stay yours — stored encrypted, never logged
regions
choose where instances run, including canadian data residency
export
download backups, export data, move plans, delete — no lock-in
state 05 running_health_checks
Designed to keep running.
Every template ships with defined health checks, backups, version controls, and recovery procedures. This is the core of QIOI, not a feature list.
container health
defined per template · auto-restart on failure
backup schedule
encrypted · restorable · verified
version pinning
stable + lts channels · staged rollouts
rollback path
failed updates return to last known-good
resource monitoring
cpu · ram · storage · usage alerts
status visibility
public status page · incident history
state 06 ready
Service ready. Here is what you can run.
spaceone ── physical compute + data centres
└─ qioi ── managed deployment + operations ◄ you are here
└─ deepfryr ── trusted knowledge + rag
└─ vecondo ── vertical applications
deployment complete — credentials