If you're using ATHENA please up token to max or custom. The actual cost won't change since it's per response, but the template is beefy and will require more tokens to use. This works very well with MX or Ame for cheaper options aswell :)