• the same as auto assigning the voice on the Character level feature recently added, having an option to change the model settings for the AI Character would allow a ton of additional use cases and also help with token usage.
  • at the moment I have an AI Character that is designed for a very specific task and if the context window is too long then it runs up the token count unnecessarily. I would like to limit this character to as few messages as possible to preserve the token usage and prevent the end user from running up a high bill.