Cheap at $0.18 an hour, it suits developers transcribing meetings with many speakers and mixed languages.
Kenning's take
What it does
Muse Voice Transcribe is Meta's first real-time audio model, turning live or recorded speech into text. It separates more than 20 speakers, detects when someone has finished talking, handles 25 or more languages including switching mid-sentence, and can be steered with keywords and context. It is available through the Meta Model API, in Meta AI for Mac and in Muse Code.
- Real-time streaming transcription
- Speaker labelling for 20+ speakers
- 25+ languages with code-switching
- Endpointing and voice activity detection
- Keyword and context biasing
Who it suits. Developers building voice agents, meeting notes, captioning or dictation products.
Key facts
- Category
- AI models
- Part of
- Meta AI
- Made by
- Meta (Meta Superintelligence Labs)
- Released
- 1 September 2026
- How to get it
- through the Meta Model API (muse-voice-transcribe-1.0), Meta AI for Mac and Muse Code
- Works on
- API, macOS
- Free plan
- platform free-tier credits apply (amount not stated)
- API price
- $0.18 per hour of audio processed
FAQ
How much does Muse Voice Transcribe cost?
$0.18 per hour of audio on the Meta Model API, the same for streaming and file transcription.
How many speakers can Muse Voice Transcribe tell apart?
Meta says it handles diarization with more than 20 speakers.
When was Muse Voice Transcribe released?
Meta introduced it on 1 September 2026.
Facts from the maker's own pages and the sources listed. Checked 2026-10-11.
Similar tools
Meta Llama models
Meta's family of downloadable large language models, from Llama 2 to the multimodal Llama 4
Muse Glimmer
Meta's open-weight 30B model for running AI agents locally on a Mac or PC with one consumer GPU
Muse Image
Meta's in-house image model that plans a layout before drawing, used in Meta AI and the Meta Model API
Muse Spark
The first model from Meta Superintelligence Labs, a multimodal reasoning model behind the Meta AI app