Voice Cloning Studio
Fish Audio Voice Cloning · Recording Booth Edition
Dark
♫
How this connects to Voice Atelier:
once a clone finishes here, press
Copy ID
on its Voice ID. Open
Voice Atelier
and paste that ID into the “Have a voice ID already?” field to generate speech in the cloned voice.
Create Model
Get Model
Delete Model
Update Model
Create Model
Title
Description
Visibility
Private
Unlisted
Public
Private is recommended unless you intend to publish this voice to the Fish Audio catalog.
Voice samples
Upload files
Record from mic
Drop audio files or click to browse
WAV, MP3, M4A, or Opus · up to 20 clips · 25 MB each
00:00
/ 03:00
Ready to record
Click to start recording a take
Use this take
Discard & retry
This browser doesn’t support microphone recording. Use the Upload files option instead.
Clean, single-speaker audio gives the best clone. A short clip works for a quick test; a minute or two of clear speech improves fidelity. Recorded takes are saved as WAV automatically.
Transcripts (optional)
If provided, the number of lines must exactly match the number of voice samples above.
Cover image
Drop an image or click to browse
JPG, PNG, or WEBP · up to 5 MB · required only for public voices
Tags
Enhance audio quality (recommended for noisy or lower-quality recordings)
Generate a default sample after training
Press the voice
Get Model
Voice ID
Fetch model
Delete Model
Voice ID
This permanently removes the voice model from your account. It cannot be undone.
Delete permanently
Update Model
Voice ID
New title
New description
New tags
Cover image updates aren’t supported here — re-create the model under Create Model if you need a new cover.
Save changes
Voice
Cloning