OmniVoice
OmniVoicev2026CurrentReleased March 31, 2026
Context--
Training CutoffNot publicly specified
CategoryAudio Model
About OmniVoice
OmniVoice is an open-source massively multilingual zero-shot text-to-speech and voice-cloning model from K2-FSA. The project README says it supports more than 600 languages, voice design controls, and fast inference with real-time factor as low as 0.025.
Capabilities
text to speechvoice cloningmultilingual audiovoice design
Input Modalities
textaudio
Output Modalities
audio
Technical Details
- API Identifier
- k2-fsa/OmniVoice
- Category
- Audio Model
Tags
text-to-speechvoice-cloningmultilingualspeech-aiopen-sourcediffusion-model
Run this model yourself
Local and cloud setup for OmniVoice, with exact configurations, sources and clear evidence limits.
No reviewed setup guide is available for this model yet. Explore configurations with documented support in the hardware finder.
Benchmarks
Performance scores for OmniVoice across standard benchmarks.
Real-time factor (best-case author claim)official-github-readme · Sep 2026
0 ratio
Pricing and access
Open-source model repository; hosted API pricing is not listed in the GitHub README.
K2-FSAopen source
Open-source speech AI and finite-state automata projects
1 ModelsFounded 2020Not publicly listed
View full profile