Yes, it is possible to include ambient background noise merged with TTS output in LiveKit. You can use the

BackgroundAudioPlayer

class to achieve this. This allows you to add realism by playing ambient background audio alongside the synthesized speech. Here is an example of how to implement it:

from livekit.agents import BackgroundAudioPlayer, AudioConfig, BuiltinAudioClip

async def entrypoint(ctx: agents.JobContext):
    await ctx.connect()

    session = AgentSession(
        # ... your stt, llm, tts, vad, turn_detection, etc.
    )

    await session.start(
        room=ctx.room,
        # ... agent, etc.
    )
    
    background_audio = BackgroundAudioPlayer(
        # play office ambience sound looping in the background
        ambient_sound=AudioConfig(BuiltinAudioClip.OFFICE_AMBIENCE, volume=0.8),
        # play keyboard typing sound when the agent is thinking
        thinking_sound=[
            AudioConfig(BuiltinAudioClip.KEYBOARD_TYPING, volume=0.8),
            AudioConfig(BuiltinAudioClip.KEYBOARD_TYPING2, volume=0.7),
        ],
    )
    
    await background_audio.start(room=ctx.room, agent_session=session)

The

BackgroundAudioPlayer

class supports two types of audio: • Ambient sound: A looping audio file that plays continuously in the background • Thinking sound: Audio that plays while the agent is processing/thinking You can adjust the volume of these sounds using the

volume

parameter in the

AudioConfig

object, with values between 0 and 1.0. Sources: Agent speech and audio | LiveKit Docs | GitHub::livekit::agents | examples/voice_agents/background_audio.py