AI

LLM Provider

Pick a hosted API or a self-hosted local server. The app runs fine with nothing configured — the AI features just stay inert until you set this up.

Running in a container? localhost won't reach a server on the host — use the host's LAN IP and bind the LLM server to 0.0.0.0. The Test button below will show the real error if it can't connect.

AI energy insights general

Sends your recent usage summary to the general model.

Identify appliance vision

Upload a photo; the vision model guesses the appliance & watts.

AI request log
WhenProviderEndpointModelTokens tok/sDurationStatusQuery
No requests yet.