Microsoft Unveils VibeVoice for Longer Conversational AI Audio
SMRTR summary
Microsoft has released VibeVoice, an open-source AI model that generates up to 90 minutes of podcast-quality speech with four distinct voices. The system uses 1.5 billion parameters and Alibaba's Qwen2.5 language model to create natural conversations, while incorporating safety features like AI disclaimers and hidden watermarks to prevent misuse such as impersonation or disinformation.
SMRTR provides this summary for quick context. The original article belongs to PYMNTS.
Read the original article