Unmute
A voice interface that connects text language models to Kyutai streaming speech recognition and speech synthesis for spoken conversations.
About Unmute
Adding spoken interaction to a text language model requires coordinating transcription, model responses and speech playback.
Who it’s for
- Developers adding voice conversations to text language models
- Teams deploying their own speech interface
When to consider it
Consider Unmute when you want to connect an OpenAI-compatible language model server to a self-hosted voice interface.
Tradeoffs & limitations
- MIT licensed. Self-hosting requires an x86_64 machine with a CUDA-capable GPU and at least 16 GB of VRAM, running Linux or Windows with WSL; native Windows and macOS are not supported.
SOTA overview · Documentation-based assessment · Sources & review method
Updates
No updates shared yet.
Discussion
Newest firstAsk a question or share how you use Unmute.
Keep it helpful. Community rules
Loading discussion…
Sources & review method
Documentation-based assessment · Oct 30, 2026 · Prepared with AI assistance; not a hands-on benchmark.
- Official voice demo
- Official architecture, model integration and hardware requirements
- MIT code licence
- Kyutai product overview and ownership
Checked by SOTA · AI-assisted documentation review. Selection advice is our assessment; verify current requirements for your deployment.
Import history & original evidence
Unmute official website
A voice interface that connects text language models to Kyutai streaming speech recognition and speech synthesis for spoken conversations.
Based on official pages and announcements checked on 2026-10-30. No hands-on test, performance benchmark or popularity ranking is claimed.