MicrosoftMAI Voice
MAI-Voice-2.1-Flash
MAI-Voice-2.1-Flash is the public-preview variant for latency-sensitive speech generation.
Preview
What the release provides
It supports 23 languages, the Azure Speech SDK and Voice Live integration. SSML controls emotion and style. Instant cloning is gated; prebuilt licensed voices can be used without that custom-voice path.MAI-Voice Azure Speech documentation
How to use it well
Practical guidance. Measure the complete conversation delay, including transcription, reasoning and tool execution. Compare speech clarity during short replies and interruptions. A fast synthesis measurement does not determine how quickly the entire assistant answers a caller.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Status | Preview |
| Access | API |
| Model identifiers | MAI-Voice-2.1-Flash |