# SiliconANGLE — Microsoft targets ultra-realistic voice agents with its first streaming transcription model

- Company: SiliconANGLE (siliconangle.com)
- Announced: 2026-10-02T01:31:31+00:00
- Category: new-model
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://siliconangle.com/2026/10/01/microsoft-targets-ultra-realistic-voice-agents-with-its-first-streaming-transcription-model/
- Record: https://forck.live/items/15667-microsoft-targets-ultra-realistic-voice-agents-with-its-first-streaming
- Subject: Reporting
- Models affected: MAI-Transcribe-2-Streaming, MAI-Voice-2.1, MAI-Voice-2.1-Flash
- Pricing: MAI-Transcribe-2-Streaming: 54 cents per audio hour; MAI-Voice-2.1: $22 per million characters; MAI-Voice-2.1-Flash: $15 per million characters.

Microsoft introduced MAI-Transcribe-2-Streaming, its first streaming transcription model, alongside MAI-Voice-2.1 and MAI-Voice-2.1-Flash text-to-speech models. The streaming model accepts speech via WebSocket, returns continuously updated transcripts, and costs 54 cents per audio hour. The text-to-speech models are priced at $22 and $15 per million characters respectively.

## Evidence

Verbatim from https://siliconangle.com/2026/10/01/microsoft-targets-ultra-realistic-voice-agents-with-its-first-streaming-transcription-model/:

> MAI-Transcribe-2-Streaming is listed on Microsoft's Vercel AI Gateway and priced at 54 cents per audio hour, the company said. It supports more than 60 languages, and can detect which one someone is speaking automatically.

---

Record: https://forck.live/items/15667-microsoft-targets-ultra-realistic-voice-agents-with-its-first-streaming
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md