2 questions: Assist history? OpenAI tokens increased, still answer cuts off too soon (piper issue, not OpenAI I guess)

All, I sooo love the new voice capabilities! I setup 2 pipelines.

One for all Hass automation and one to talk with OpenAI/ChatGPT. I have 2 questions. I guess I can individually trigger them by assigning different wake words.

Too bad on my WearOS watch it only works with the favorited pipeline.

  • Is there somewhere in Hass where I can find the conversation/chat history I have had with both of the pipelines? On a very quick search I couldn’t find anything.

  • I sort of understand the ChatGPT API tokens stuff. So I increased the number of tokens that are allowed in a ChatGPT response. However the Speech response is still cut off very early. See screenshot response. The speech will stop reading after speaking the first bullet “1.” I take this is not a token issue, since I got the full OpenAI response in text. Why is Piper then cutting off the spoken text?