The idea is:
Expose Full Chat Model Output in AI Agent Node. Add an option in the AI Agent node (e.g., a checkbox like “Return full response”) that outputs all data received from the Chat Model API, similar to what is available when using the Chat Model node directly.
My use case:
I think it would be beneficial to add this because:
Currently, the output of the Chat Model inside the AI Agent node is not fully accessible in subsequent nodes. Only the model’s text response is available, while other useful data (like token usage, cost, model details, or raw API response fields) is missing.