To display a transcript of a realtime interaction in your bootstrapped voice assistant frontend, LiveKit's Agents framework supports delivering realtime transcriptions during voice interactions. Here's how it works: • User speech transcriptions are delivered as soon as they are processed by the Speech-to-Text (STT) system. • Agent text responses are synchronized with audio playback. If you're using

VoicePipelineAgent

or

MultimodalAgent

, transcriptions are automatically forwarded to clients when you implement the necessary frontend support. For custom agents, you can use the

STTSegmentsForwarder

class to forward transcriptions to your frontend in realtime. Sources: AI voice agents | LiveKit Docs | Transcriptions | LiveKit Docs