I am looking for an open source (i.e., free) Text-To-Speech library that works offline (i.e., data need not be sent to cloud like in case of Google Cloud Speech service) and can be sold (i.e., is commercially available, unlike some GNU licences) and is in any programming language for any Operation System.