Pricing Built the Way a CFO Would Want It

Enterprise buyers think in two lines: capex or a project line for the build, opex or a subscription line for the run. Aegis Private AI pricing matches that exactly. Part one is a one-time professional services engagement at each build stage. Part two is a recurring managed subscription that runs continuously once you are live. You are paying IVI to design and build the environment, and then to operate it, with the compute cost transparently passed through. Nothing is hidden inside a blended per-token number.

The recurring fee has two components, deliberately separated. The consumption pass-through is the raw GPU infrastructure cost, disclosed openly with a modest 10 to 15 percent markup. The managed services tier fee is the profit driver, and it is priced against the alternative of hiring and running your own ML platform team, not against the GPU cost. That matters: when GPU prices move, your pass-through line moves and your tier fee does not. You are insulated from the compute market, and you never pay a hidden margin on hardware.

The subscription runs continuously across the whole journey. In Stage 1 the pass-through line is rented GPU-hours. In Stage 3, after you own the cluster, it becomes amortized on-prem cost, and the tier fee scales with cluster size. What does not change is the operational relationship: one continuous managed service, the same team and playbook, regardless of where the infrastructure physically sits.

Private AI isn't the right call for every team, and we'd rather tell you that in the first conversation than the sixth month. Here's how we think about fit.