Skip to main content
Displayed at all times in the lower-right corner of the application, the resource consumption metrics provide real-time information about your system’s CPU and RAM usage, both overall and per loaded model. If no model is loaded into the API server or chat interface, the metrics display No Models Loaded. Loading a model into the API server or the chat interface updates the display to show you how much CPU and RAM the loaded model is currently consuming. Hover over the metrics to display:
  • Additional information about how many models are currently loaded
  • How much system CPU and RAM you have
  • How much is being consumed by the models
  • How much is dedicated to other processes (outside of Desktop)
Click Loaded Models in the status bar to open the resource consumption pane. The pane has four sections:
  • Chat: Shows any model currently loaded in the chat interface. Click View Chat to open the chat. Click Eject to unload the model and free up consumed resources.
  • Server: Shows any model currently loaded in a running model server. Click View Server to open the Model Servers page. Click Eject to stop the server and free up consumed resources.
  • Other System Processes: Shows your total system CPU and RAM usage.
  • Storage: Shows disk usage broken down by application and downloaded models.
Resource consumption pane showing loaded models with CPU and RAM metrics