Common questions
Is Aurius the same thing as the acelle/ai plugin?
The acelle/ai plugin is the UI — chatbox, sparkle, coaches, admin dashboards. Aurius 4.0 is the managed AI service the plugin can call into. If you already have an OpenAI or Anthropic key, the plugin works fine on its own; Aurius is the alternative for teams that don't want to manage keys or balances.
Can I bring my own AI engine?
Yes. The plugin supports OpenAI, Anthropic, and self-hosted Ollama natively. Add the engine, test the connection, set it as default — that engine handles every chatbox, sparkle, and coach call. Aurius isn't a lock-in.
What does “a credit” mean?
Aurius bills by usage. A short chatbox turn or one sparkle rewrite is a small amount of credit; a long agent loop that fans out to multiple tools costs more. Every plan publishes the per-task averages so you can predict where you'll land — and the admin Usage dashboard shows real-time spend.
Where does my data go?
Every prompt, response, tool call, and feedback rating lives in your Acelle database — eight audit tables prefixed ai_*. Aurius itself sees only the request payload needed to generate the response; nothing is used for training. See our privacy policy for details.
Can I disable AI for individual users or screens?
Yes. Aurius exposes a per-feature flag matrix (Support chatbox, Agent tool-use, Sparkle rewrite, Coach personas, Admin AI dashboards) and a Permissions matrix — admin only, all users, none. Flip it from AI Settings and the UI surfaces hide instantly.
Does Aurius work with the Extended (SaaS) license?
Yes. Multi-tenant Acelle installs running the Extended license can run Aurius across all tenants from a single key, and the included Permissions matrix gates which subscription tiers see which AI surfaces.