For some workloads the question is not which model performs best. It is whether sending the data anywhere is permitted at all — and that decides whether the project proceeds.
Asked to approve an AI workload that would send regulated data to a third party.
Data residency and contractual restrictions constrain where inference may run.
Connectivity is intermittent or absent, so a hosted call is not an option.
| Approach | What it requires | Where it falls short |
|---|---|---|
| Hosted provider + DPA | Legal review, vendor assurance | Data still leaves; some rules disallow it outright |
| Self-host an open-weight model | GPU hardware, serving stack, ops ownership | Real infrastructure cost and a system to run |
| Do without | Nothing | The capability is simply not delivered |
| NC AI | Install one binary | Smaller model — see limitations below |
NC AI is the inference engine inside the NC runtime. There is no serving stack to deploy, no GPU requirement, and no account to create.
The constraint is regulatory, contractual or connectivity-driven; the task is bounded and structured; cost predictability matters more than peak capability.
Open-ended generation, long-context reasoning, or anything where output quality is the product. Use a hosted model.
Install it, disconnect the network, and confirm it still runs. That is the claim.