Modern voice cloning (ElevenLabs, OpenAI, Cartesia) can produce a clone from 10 seconds of source audio. Quality is high enough for most brand-voice use cases.
Ethical voice cloning requires explicit consent from the voice donor and clear disclosure to listeners. Callsy supports brand-voice cloning on Pro and Scale plans with signed consent.
Cloned voices outperform stock voices by 5–15% on brand-recall metrics, but the absolute conversion lift is usually marginal. Script quality matters more than voice identity.