Effective date: 1 June 2026
Version: 2026-06-01
SovereignEG provides an OpenAI-compatible API that may route requests through two types of backing infrastructure:
- Standard provider routing: requests may be sent to third-party model providers when a selected model is backed by a serverless provider account. Those providers process the request payload needed to generate the response.
- Dedicated / Egypt-hosted paths: selected private deployments may run on dedicated infrastructure controlled by SovereignEG or the customer. Availability depends on the model and commercial setup.
Model availability, provider backing, and routing can change as we add,
remove, or disable backings. The public model catalog shows which models are
available and their region label; the "provider" it shows is usually the
company that created the model, not the company that serves your request, and
it does not name the third-party provider serving a model. For models served by
third-party providers, signed-in customers can see the current providers, in
priority order, in the engines field of GET /v1/models and on the model's
page in the dashboard.