Skip to main content

System AI overview

System AI is a shared AI capability that Enhanced provides on top of the open-source core. Open source has no system-level AI; Enhanced centralizes models, credentials, and billing so projects and the Studio assistant can use them directly.

What users can do​

  • Open Settings → AI settings to choose a default model, reorder provider preference, subscribe to a plan, or use balance.
  • Check requests, tokens, cost, and subscription records under Usage and Orders.
  • Use System AI directly in the Studio assistant and projects without configuring a model key.

How a request flows​

  1. A user or project asks for an available model.
  2. The Server returns the models from groups the admin has opened and issues a short-lived credential.
  3. The request goes through the Server's AI gateway to the model service.
  4. The platform meters it against the plan or balance and records usage.

Availability and defaults​

  • Only groups the admin has opened and that are currently usable appear in personal AI settings.
  • Default-model precedence: personal default → group default → project built-in default.
  • Provider preference only orders available groups; an explicit provider choice is never overridden.

Billing and quotas​

  • Subscription plans run in 30-day periods; balance is shared at the account level.
  • Usage and cost are recorded by the platform; quotas come from the plan or balance.

Getting started​

  1. Open Settings → AI settings.
  2. If the page is empty, ask an admin to open a group or bind your account.
  3. Choose a default model and subscribe or recharge as needed.
  4. Open the Studio assistant to verify the connection.

Continue reading​