FAQ

How do I use a model saved in API Connection Settings from an Agent Profile?

An Agent Profile does not store API keys or one-off model input directly. It references a model configuration that you have already saved in API Connection Settings.

Use this order:

1. Open API Connection Settings.
2. Add or select a usable model, then save it. Make sure to choose Save model only.
3. Open the Agent System panel.
4. Open the Agent Profile you want to use.
5. In the model binding area, choose Bind model connection or the equivalent option.
6. Select the model you saved.
7. Save the Agent Profile.
8. Return to the Agent panel and confirm that Active Profile is the profile you want to run.

If an imported Profile says it requires configuration, it usually is not broken. Exported Profiles do not carry someone else's local model connection. Re-select a model on your machine and save the Profile.

If the model does not appear in the Agent Profile picker, check these points first:

• API Connection Settings was saved, rather than only filled in.
• The model uses a chat-completion path supported by the current Agent system.
• You reopened or refreshed the Agent configuration panel after saving the model.
• The Profile you are editing has been saved.

If you only want Agent to use the same model as the current chat, keep Current prompt snapshot. You only need to bind a model in the Profile when you want that Agent to use a specific saved model.