Voice Library
Manage cloned voice models with automatic voiceprint matching
Cloned Voices
No cloned voices yet
Upload a video and enable "Keep Original Voice" to automatically create a voice clone
Voiceprint Match Test
Upload an audio file to test whether it matches an existing voice in the library. Similarity >= 0.80 will be reused automatically; 0.65-0.80 requires manual confirmation.
Click or drag an audio file here
Supports WAV, MP3 formats
Voice Clone Persistence Mechanism
P1: Exact matching via audio file hash — identical reference audio directly reuses the cloned voice model.
P2: Fuzzy matching via AI voiceprint recognition — different recordings of the same speaker can also be reused automatically.
Matching thresholds: similarity >= 0.80 auto-reuse, 0.65-0.80 requires confirmation, < 0.65 creates a new clone.
P1: Exact matching via audio file hash — identical reference audio directly reuses the cloned voice model.
P2: Fuzzy matching via AI voiceprint recognition — different recordings of the same speaker can also be reused automatically.
Matching thresholds: similarity >= 0.80 auto-reuse, 0.65-0.80 requires confirmation, < 0.65 creates a new clone.