nvidia / PersonaPlex-Realtime
PersonaPlex-Realtime
Convert speech to speech with customizable voices using PersonaPlex. Supports various audio formats and offers control over text temperature and audio top K settings.
## Overview PersonaPlex is a tool designed for converting speech to speech with customizable voice options. It supports multiple audio formats and allows fine-tuning of output using advanced parameters such as text temperature and audio top K. ## What you can build - Voice-over content with different speaker styles - Audio-to-audio transformations for multimedia projects - Customizable voice synthesis for digital assistants ## Inputs - **Input Audio**: Supported formats include .wav, .mp3, .ogg, and .m4a - **Prompt**: Text prompt to guide the response - **Voice**: Selection of predefined voice options (Natural or Variety) ## Outputs - Transformed audio with specified voice characteristics ## Recommended settings - Use default values for basic functionality - Adjust text temperature for variation in generated responses - Modify audio top K for finer control over audio generation ## Limitations - Input audio must be clear and compatible with supported formats - Voice selection is limited to predefined options ## Safety & compliance - Ensure input audio complies with copyright regulations - Use generated outputs responsibly and in accordance with applicable laws
Example prompts
Great starting points for PersonaPlex-Realtime.
API quick start
Run PersonaPlex-Realtime with a single API call.
{
"prompt": "You are a helpful assistant.",
"voice": "VARF2",
"tempText": 0.7,
"topkAudio": 250
}