MAI-Transcribe-2
Notable release: A meaningful new model: a new size or version, or a strong free-to-download model from a known lab.
Microsoft's MAI-Transcribe-2 turns recorded speech into written text and can identify who is speaking, offered to developers online.
Key facts
- Released
- Sep 3, 2026
- What it does
- Listens: Understands speech and other sounds.Makes sound: Creates speech, music or other audio.
View moreView less
- How
- For developers, through an API
- Announced as
- “MAI-Transcribe-2”
From trckr's archive of releases before it started watching. Date and details from Epoch AI's database of AI models (CC BY 4.0).
Compare MAI-Transcribe-2 with
It came out 42 days after MAI-Voice-2-Flash, the lab's previous model.
The details
MAI-Transcribe-2 is Microsoft's audio model for Speech-to-text: Turns spoken audio into written text. Also called transcription or ASR. and voice identification, accessed through an API: How developers plug a model into their own apps and pay per use. A model that's "in the API" can be used by programmers, even if it isn't in a chat app yet. and not released as a download.
Get alerts like this the moment they happen.
From Microsoft and every other major lab.
Free, no account needed. Push, Telegram, Discord, email or RSS.