Frameworks
LiveKit
Build LiveKit voice agents with Prism speech-to-speech or Rime text-to-speech. Choose a model, install the plugin, and connect your agent.
Use Prism as your agent’s speech-to-speech model, or use Rime text-to-speech with your own transcription and language model. LiveKit Agents manages the room and agent session in both cases.
For Prism, follow Build a LiveKit voice agent with Prism to install the plugin and connect your agent.
The Prism LiveKit plugin reference covers settings, tool calls, retries, playback, and conversation history.
For text-to-speech, build a complete agent with the LiveKit TTS quickstart, or use LiveKit’s Rime TTS reference for plugin configuration.
For Coda WebSocket v1, use the LiveKit WebSocket integration for agent configuration and plugin settings.

