OpenVoice
Open-source voice cloning uses a reference recording to generate multilingual speech with style control. The Python project uses the MIT license.
OpenVoice is an open-source voice cloning tool that uses a short recording to reproduce a speaker's voice in generated speech. It's for developers and creators who need a recognizable voice across languages, with control over how that voice sounds. The Python project is MIT licensed for commercial use.
Voice identity and delivery can be controlled separately. OpenVoice reproduces the speaker's vocal tone while allowing changes to emotion, accent, rhythm, pauses and intonation. That gives users room to vary the performance without choosing a different reference speaker. Demonstrated styles include happy and sad speech, as well as British, Indian and Australian accents.
Its cross-language cloning doesn't require the reference recording and generated speech to share a language. It can also clone voices across languages absent from its multilingual speaker training data. OpenVoice V2 natively supports English, Spanish, French, Chinese, Japanese and Korean. This makes it relevant to multilingual narration and speech projects that need to carry one speaker's voice into another language.
OpenVoice's approach emphasizes computational efficiency alongside voice and style control. Developers can use its source code in their own projects under the MIT license. MyShell also uses the model to provide instant voice cloning on its hosted platform.
