Microsoft Azure AI Voice Live API

The Microsoft Azure AI Voice Live API, launched June 2025, is Microsoft's enterprise-grade orchestration layer for building real-time AI voice agents. It combines Azure AI Speech STT across 50-plus locales, Azure OpenAI LLM, and over 600 premium neural TTS voices spanning 150-plus locales in a single billable endpoint, with response latency under 500 milliseconds. The API supports multilingual STT across 15 locales simultaneously and offers Pro and Basic pricing tiers with per-token audio billing. BYO-LLM and BYO-STT configurations are supported for organizations that need to substitute fine-tuned models. Custom Neural Voice training integrates directly for brand voice deployments, and Azure Communication Services provides native SIP trunking and cloud telephony connectivity to replace legacy IVR systems. Azure AI Voice Live API targets enterprise contact centers requiring 99.9 percent uptime SLAs, compliance certifications, and seamless integration into existing Azure infrastructure. Organizations already on Azure benefit from unified billing, managed identity security, and direct connections to Azure Contact Center AI and Dynamics 365 Customer Service for end-to-end CX deployments.