astonishing-ability-75576
08/04/2025, 8:57 PMtall-airline-16756
08/04/2025, 8:57 PMtall-airline-16756
08/04/2025, 8:58 PM.env file as DEEPGRAM_API_KEY.
2. Install the Deepgram Plugin: Run pip install "livekit-agents[deepgram]~=1.0".
3. Initialize Deepgram STT: Create a function to configure the STT module with options like interim results, punctuation, and language.
4. Configure the Agent: Define an agent class to handle transcription and log events.
5. Set Up AgentSession: Use AgentSession.start to include both SIP and web participants by specifying rtc.ParticipantKind.PARTICIPANT_KIND_SIP and rtc.ParticipantKind.PARTICIPANT_KIND_STANDARD in RoomInputOptions.
6. Handle Multiple Participants: Implement logic to start a session for each participant dynamically as they join the room.
For detailed code examples and additional considerations, refer to the Deepgram STT Integration Guide and LiveKit Agents Framework Documentation.astonishing-ability-75576
08/04/2025, 9:07 PMtall-airline-16756
08/04/2025, 9:08 PMimport { initializeLogger } from '@livekit/agents';
initializeLogger({
pretty: false,
level: 'debug'
});
2. Using `setLogExtension`: This lets you send logs to external providers like Datadog, which is super handy. import { setLogExtension } from '@livekit/agents';
setLogExtension((level, msg, context) => {
// Just plug in your log provider's API here to send this log message and context
});
These options will get you detailed debug logs for the STT module in no time. If you have any questions or want a hand with anything else, just let me know—I’m here to help!
Sources: Logs, metrics, and telemetry | LiveKit Docs | Voice Assistants | livekit/livekit_composite | DeepWiki