For my use-case, I want the providers to get my tokens as long as they plan to keep releasing open-weight models