How you run Hive Three tiers Canada-led

One Hive. Three ways to run it. You choose how much control you keep.

From a box you own and never let out of the building, to a region-locked shared instance you can start in a day, the same Hive software runs across the whole spectrum. You pick the point that matches your data and your budget.

One stack, no rewrites

The software does not change. Only where it runs does.

Whichever way you run Hive, the underlying system is the same OpenAI-compatible gateway, with the same models, the same API surface, and the same admin console. Your applications do not change as you move along the spectrum. You can start on a shared instance to prove value, move to a dedicated colocated box as your data sensitivity grows, and bring the box fully on-prem when control becomes non-negotiable, without rebuilding your integration.

The one rule across every hosted tier

No hosted tier runs outside your own jurisdiction.

This is the rule that makes the hosted tiers worth offering. The reason on-prem control matters is that any cloud provider is reachable under the law of the jurisdiction it answers to, wherever its datacentres sit. If we then hosted you under a jurisdiction other than your own, we would hand back the exact exposure you came to us to remove. So every hosted Hive tier runs only on infrastructure that is in-region and under your own jurisdiction.

An honest boundary

A managed tier on any third-party infrastructure does not carry the same data-residency properties as the on-prem edition. If your hard requirement is that data stays on hardware you physically control, the on-prem tier is the one that meets it.

A spectrum of control

The same Hive, from maximum control to lowest cost.

Three tiers sit on one continuum. Isolation is highest on the left and the operational burden you carry falls as you move right. Every tier runs the same software. The three sections below describe each tier in depth, and we grade each one honestly.

  • 01
    Own it On-prem EnterpriseEdge
    Sovereignty
    Highest
    Ops burden
    Your team, with our support
  • 02
    Colocate it Dedicated box, hosted by us
    Sovereignty
    High, third-party hosted
    Ops burden
    Low, we manage it
  • 03
    Shared cloud Region-locked, fully audited
    Sovereignty
    Lowest, region-locked
    Ops burden
    None, we run everything
Maximum control

Own it. A Hive deployment on your premises, run by your team.

EnterpriseEdge is the full Hive stack running on a box that lives in your building, owned and operated by you. It starts at a single workstation-class box for a small regulated firm and scales to multi-GPU and multi-device as load grows. It is the same software the hosted tiers run, entirely inside your walls. Air-gapped and offline deployments are supported for environments that permit no outbound network path.

Who it suits
Teams whose data cannot leave the premises under any circumstances. Financial institutions handling material non-public information, healthcare custodians holding personal health information, government and defence programmes, and law firms protecting privilege.
Sovereignty level
Highest. On the on-prem deployment, data never leaves your network boundary and no third party is in the data path. This is the only tier where the full sovereignty claim holds.
Ops burden
You run it. The box, the updates, and the hardware sit with your team, backed by our support and your annual licence. Highest control, most operational ownership.
Regions
Anywhere you can place a box. There is no datacentre and no region question, because the hardware is yours.
Dedicated and in-region

Colocate it. Your dedicated box, hosted by us in your region.

We host a Hive box dedicated to you, in-region, on infrastructure under your own jurisdiction. The hardware is not shared with any other customer. We manage the machine so your team does not have to, and residency stays in-region.

Who it suits
Teams that need dedicated, in-region hardware and strong residency, but do not want to run a box themselves. A middle ground when on-prem operations are more than the team wants to carry.
Sovereignty level
High, with one honest caveat. The box is dedicated and in-region, under your own jurisdiction, which removes the cross-jurisdiction reach that an out-of-region host would carry. It is still hosted by a third party rather than physically held by you, so it does not match the on-prem tier byte for byte.
Ops burden
Low. We manage the hardware, updates, and uptime. You get a dedicated box without running one.
Regions
Canada is the confirmed direction, hosted with an in-region Canadian provider or research-cloud option under Canadian jurisdiction. We also serve the UK and Bangladesh, with in-region providers being confirmed. We will not name a facility until it is locked.
Lowest cost, fastest start

Shared cloud. Region-locked, fully audited, honestly the lowest tier.

A shared Hive instance, region-locked, with a full audit trail of every request. It runs only on in-region infrastructure under your own jurisdiction, the same rule as the colocated tier, but capacity is shared across customers rather than dedicated to you.

Who it suits
Teams that are comfortable sharing capacity and want the lowest cost and the fastest start. A practical way to prove value before committing to dedicated or on-prem hardware.
Sovereignty level
Lowest of the three, and we will not dress it up. It is region-locked, fully auditable, and runs on infrastructure under your own jurisdiction, but it is not physically isolated and it is not on hardware you control. If physical isolation or on-prem control is your requirement, this is not the tier for you.
Ops burden
None. We run everything. You point your application at it and go.
Regions
Region-locked to the region you select, on in-region infrastructure under your own jurisdiction.
Capacity you add, not a box you replace

When a workload outgrows a box, you add to it.

Hardware needs change as models grow and usage climbs. Hive is built to extend rather than to be thrown out. You start on a single box sized to today's workload. When throughput needs more, you add GPU capacity to that box. When one box is not enough, you scale to multiple nodes. The integration does not change as you grow, so the investment compounds instead of resetting. A box bought today stays useful as the platform and the models advance.

Under the annual licence

Your deployment keeps pace with the field.

Owning the hardware does not freeze the software. The annual licence keeps the Hive platform current and adds compatibility with new open-weight models as they ship, so your box runs newer and stronger models over time without a new purchase. You are not buying a fixed snapshot. You are buying a platform that keeps improving on hardware you already own.

Side by side

The spectrum at a glance.

The three Hive delivery tiers compared across six dimensions.
Dimension Own it (on-prem) Colocate it Shared cloud
What runs where Your box, your building Your dedicated box, our in-region datacentre Shared instance, region-locked
Sovereignty Highest High, third-party hosted Lowest, region-locked and audited
Data isolation Physical, on your premises Dedicated hardware Shared capacity
Ops burden Your team, with our support We manage it We manage it
Infrastructure Yours In-region, your jurisdiction In-region, region-locked
Best for Data that cannot leave the building Dedicated and in-region without running a box Lowest cost and fastest start

Sovereignty grading describes architectural posture, not a compliance determination. Validate fit for your obligations with your own counsel.

Not sure which tier fits? We will help you place it.

Tell us about your data and your obligations, and we will recommend the point on the spectrum that fits.