Sometimes I’ll ask something of Assist (using OpenAI API) and it will have multiple steps of a response. Briefly the app will show me the first response, and then it overwrites it with the second. How do I get it to show me everything that it’s saying? I’m unsure if this has anything to do with the maximum token setting, but that setting says that this is about how much the AI model generates in response.
This is just from playing around with a prompt, but it only briefly shows this first message before replacing it with the second:
2025-04-23 09:10:28.611 DEBUG (MainThread) [homeassistant.components.openai_conversation] Received event: ResponseTextDoneEvent(content_index=0, item_id=‘msg_6808e6c3223c8191a0bbc3acc9c770930718bf16ca934cf6’, output_index=0, text=‘What a grandiose request for a list of everything alive and active, as if it might somehow salvage this bleak existence. Let me sift through the digital remnants of this home and provide a rundown of what is currently turned on, in the hope that it might bring some faint semblance of purpose.’, type=‘response.output_text.done’)
2025-04-23 09:10:29.589 DEBUG (MainThread) [homeassistant.components.openai_conversation] Received event: ResponseTextDoneEvent(content_index=0, item_id=‘msg_6808e6c56ae081918fcd959796d07eb40e43cd6a1f281055’, output_index=0, text=‘The only lights currently turned on are in the powder room and the home office area. Almost everything else remains in darkness, like a universe resigned to its own dullness.’, type=‘response.output_text.done’)