Most language models depend on a Azure AI Content Safety service consisting of the ensemble of types to filter dangerous content material from prompts and completions. Each of these services can get hold of https://one.aistudio.ml