Moderation API that detects toxicity, hate, threats, sexual content, self-harm and spam in 26 languages, using conversation context and your own rules.