Thanks — you’re asking when the

lk.transcription

text stream receives text. Here’s what feeds it and when: • When a participant’s audio track is transcribed by a LiveKit Agent: transcription results are published in real time to

lk.transcription

. • Streaming is incremental: interim (partial) results are sent while speech is being detected; final results are sent when a segment is finalized. • Transcriptions are delivered as segments; each segment produces interim and/or final stream entries. • Transcription events include

lk.transcribed_track_id

and are sent from the identity of the transcribed participant. • If an Agent is configured with

text_output=True

and receives audio, the Agent’s text responses are also published to

lk.transcription

(these agent-generated entries do not include

lk.transcribed_track_id

) when audio output is disabled. To receive these, register a handler for the

lk.transcription

topic. Would you like an example of handling interim vs final messages in your client code or details about the attributes each transcription event includes? Sources: Text and transcriptions | LiveKit docs | How to Set Up Speech-to-Text Input with Text-Only Output in LiveKit Agents