billing qioi estimate
A bill you can read in one pass.
Hosting and management in one line item. Model API usage stays with your own provider, under your own keys. Prices are introductory and resource-based — current deployments keep their rate with notice.
For individual applications and experiments
$39/mo
per instance / month, from
For active business use
$89/mo
per instance / month, from
For larger or sensitive workloads
scoped
scoped per deployment
rate card — instance plans
| plan | vcpu | ram | storage | transfer | backup | $/mo |
|---|---|---|---|---|---|---|
| S | 2 | 4 gb | 80 gb | 4 tb | 40 gb | 39 |
| M | 4 | 8 gb | 160 gb | 8 tb | 80 gb | 69 |
| L | 8 | 16 gb | 320 gb | 12 tb | 160 gb | 119 |
| XL | 16 | 32 gb | 640 gb | 20 tb | 320 gb | 219 |
every plan includes monitoring, tested updates, backup orchestration, domain + ssl
how a qioi bill decomposes
infrastructure
- · cpu / ram / gpu
- · storage
- · bandwidth
- · backup storage
management
- · application management
- · monitoring and automatic restarts
- · tested updates and rollbacks
- · domain and ssl
- · support tier
usage (billed separately)
- · model api costs (paid to your provider)
- · document processing beyond plan limits
- · additional users
- · data transfer beyond plan limits
services (optional)
- · initial setup and migration
- · custom configuration
- · white-label configuration
- · premium support
notes
q: what happens if i exceed my transfer allowance?
We alert you before limits become incidents. Overage is billed at a published per-TB rate, or you can resize the plan — no throttling without warning.
q: can i export my data and leave?
Yes. Backups are downloadable, application data is exportable, and deletion is self-serve. Portability is a design requirement, not a favour.
q: who pays for model api usage?
You do — directly to your model provider, under your own API keys. QIOI never marks up tokens and never sees plaintext keys in logs.
q: how do updates work?
Stable and LTS channels. New upstream versions are staged and verified before rollout; a failed update rolls back to the last known-good version.