xAI has unveiled a new voice cloning feature through its API, allowing users to create production-grade voice models with just one minute of recorded speech. The system verifies voice ownership and processes audio within two minutes, supporting multilingual output and streaming via REST and WebSocket. The feature includes over 80 built-in voices across 28 languages and ensures security with a two-stage verification process. Users cannot clone voices from existing recordings or others' voices, and there are no extra charges for using custom voices with the TTS or voice agent API.