OpenVoice

A key feature of OpenVoice is the ability to adjust the style after voice cloning — it can replicate a vocal tone using just 3 seconds of audio, while allowing independent control over parameters such as tone, emotion, accent, speech speed, and pauses. It supports cross-language cloning (generating English speech from Chinese audio). The software is open-sourced under the MIT license. It is particularly suitable for various applications that require flexible adjustment of voice styles — the same vocal tone can be used to read news, tell stories, or create advertisements.