Upload a recording
Add an interview, podcast, meeting or video containing the target speaker.
Voice cloning dataset maker
Turn messy recordings into clean voice datasets.
VoiceKit finds the speaker you want, removes background sound, selects and transcribes the best clips, then exports a voice pack ready for AI voice cloning.
One recording. One clean voice.
From a long file to a usable dataset in three steps.
Choose the right voice
Preview detected speakers or add a short reference for faster matching.
Download the voice pack
Get clean WAV clips, transcripts, a best sample and a ready-to-use ZIP.