ioMoVo runs in three ways. The right one usually comes down to where your data is allowed to sit and who is allowed to hold the keys.
Cloud
ioMoVo hosts the platform and you sign in through ioMoVo's own identity layer. This is the fastest way to start, and it is the right fit when your content is not subject to residency or isolation requirements that rule out a hosted service.
You can still bring your own storage in this model. Your assets can sit in your own buckets while ioMoVo handles the application layer.
On premise
You run ioMoVo inside your own infrastructure. Your team controls the environment, the network path, and the storage. This is the common choice when a compliance regime requires the data and the processing to stay in a specific place, or when you already have significant storage investment you want to keep using.
Air gapped
ioMoVo runs entirely inside an isolated environment with no route to the public internet. Identity and governance are controlled by your own air gapped environment rather than by ioMoVo, which means your existing directory and policy stack stays authoritative.
The thing worth knowing here is that air gapped is not a stripped down version. See What works in an air gapped deployment for the detail.
Questions that usually settle it
Does your data have a residency or isolation requirement, and is it written into a contract or a regulation?
Who has to hold the encryption keys and the identity system?
Do you already own storage you need to keep using?
Does the AI processing itself need to happen inside your boundary, or only the storage?
That last one catches people out. If your policy says model inference cannot touch an external service, that points to on premise or air gapped even when your storage question alone would not.
Can you change later
Yes. Talk to your ioMoVo contact before you plan a move, because the identity path differs between models and that is the part that needs sequencing.