Before you use it
The code is open source (Apache-2.0), but the OmniVoice voice model is licensed for non-commercial use only. Check its licence before using the voices in anything that earns money, such as monetised videos.
Why it's useful
Try it free in the browser on Hugging Face, or run it yourself. Very wide language coverage, good for testing dubbing and voice-overs. Only clone voices you have permission to use.
What it needs
A graphics card (NVIDIA, Apple Silicon or Intel Arc) to run it yourself; the online demo just needs a browser. The model is licensed for non-commercial use only.
Image: k2-fsa. OmniVoice belongs to k2-fsa; we just think it's worth knowing about.
More tools like this
Open source · MITAudio & Voice
Handy
by CJ Pais
Free, open-source push-to-talk speech-to-text that runs on your own computer.
Open source · MITAudio & Voice
Unmute
by Kyutai
Kyutai's open-source kit that gives any text model ears and a voice, with interruptible back-and-forth talk.
Open source · AGPL-3.0Audio & Voice
VoiceStudio
by debpalash
An open-source, fully local alternative to ElevenLabs: voice cloning, voice design, dubbing and text to speech on your own computer.