We are excited to announce a new embedding model which is significantly more capable, cost effective, and simpler to use.