Zero-shot voice conversion using Seed-VC (v1.0 tiny · 25 M params · XLSR-300M + HIFT). Pick a source clip (the speech to re-voice) and a reference voice (the target tone) — both bundled from the ghananlpcommunity datasets — or upload your own.
Datasets used as reference voices are not reused as source clips.
1 · Source audio (the speech to re-voice)
Bundled source clip
Pre-extracted from ghananlpcommunity datasets (never the same as reference).