MODEL LINEAGE
中文
All families

FAMILY / MICROSOFT AI

MAI Speech

Microsoft AI’s speech model line for real-time transcription and multilingual speech generation.

First model
2026
Models
3
Latest dated version
MAI-Voice-2.1-Flash

01 / RELEASE SEQUENCE

Family release sequence

Compare versions →
  1. 2026-10-01MAI-Transcribe-2-StreamingLanguage
  2. 2026-10-01MAI-Voice-2.1Language · Multimodal
  3. 2026-10-01MAI-Voice-2.1-FlashLanguage · Multimodal
3 versions
What changed ↓

02 / GENERATIONS

What changed

01

MAI-Transcribe-2-Streaming

Microsoft’s first streaming transcription model, extending MAI speech recognition from post-recording transcription to real-time incremental recognition.

02

MAI-Voice-2.1

MAI-Voice-2.1 advances Microsoft’s speech generation to multilingual synthesis that preserves a consistent voice across languages.

03

MAI-Voice-2.1-Flash

Microsoft released it as a separately named Flash model alongside MAI-Voice-2.1, adding a low-latency real-time branch to the speech line; this record does not infer separate architecture or weights.