Rime has introduced Prism, a speech-to-speech AI model designed for enterprise voice agents.

The model understands emotional cues and vocal nuances, continues listening while speaking, and allows responses to be inspected before they are spoken. Rime claims Prism can react in approximately 300 milliseconds and outperformed competing real-time voice models from OpenAI and Google in its own benchmarks.