🗣️ Kanade Voice Cloning

Model: frothywater/kanade-25hz-clean

Convert any audio into a target voice. Upload a source audio (what to say) and a reference audio (whose voice to use).

⏱️ Limit: Audio is trimmed to 30 seconds max.

Tips:

• For best results, use clean reference audio (3-10 seconds of clear speech)
• Source and reference should ideally be similar in speaking pace