Microsoft Speech recognition plus MAI transcription and voice models.
8 integrated models · Official documentation ↗
Microsoft AI's newest batch speech-to-text model (September 2026, in preview on Azure), covering 60 languages with automatic language detection, code-switching, speaker diarization and word timestamps. A real-time sibling, MAI-Transcribe-2-Streaming, followed on 2026-10-01.
June 2026 update to Microsoft AI's transcription model that widened coverage to 43 languages and added phrase-list keyword biasing and clean or verbatim output styles; file input, no diarization. Superseded by MAI-Transcribe-2 in September 2026.
Microsoft AI's expressive text-to-speech model, in Azure preview since June 2026 with prebuilt voices in 15 languages and gated zero-shot voice cloning from 5-60 second clips. Superseded on 2026-10-01 by MAI-Voice-2.1 and the low-latency MAI-Voice-2.1-Flash.
Microsoft AI's first speech-to-text model, previewed in Azure Speech in April 2026 for 25 languages through the fast transcription API, file input only. Deprecated on 2026-08-20; succeeded by MAI-Transcribe-1.5 and MAI-Transcribe-2.
Lower-latency, lighter variant of Azure's Dragon HD voices (names ending DragonHDFlashLatestNeural), priced like standard neural voices. Previewed in March 2025 with Mandarin voices and extended to US English in March 2026; available in four regions.
Azure AI Speech's synchronous fast transcription API, previewed July 2024 and generally available since November 2024, returning file transcripts faster than real time with diarization of up to 35 speakers, automatic multilingual recognition and phrase lists.
Azure AI Speech's prebuilt neural text-to-speech catalog, now called standard voices: more than 500 voices across 100+ languages and locales, with full SSML control and 24 or 48 kHz output. Neural HD, HD Omni and MAI voices sit above it.
Azure AI Speech's streaming speech to text through the Speech SDK, returning interim and final results during continuous recognition, with live speaker diarization via ConversationTranscriber, phrase lists and spoken-language identification. Part of the service since its 2018 launch.
