what does livekit cli does

Thanks for reaching out, Reuban. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.

oh now i have build basic voice agent with stt,tts, llm, vad and turn detection. now i like to monitor speed, and token consumption. now what should i do next

If you want to keep an eye on speed and token usage for your voice agent, LiveKit’s got you covered with its metrics collection system. Here’s a quick rundown: • Metrics Classes: Use

LLMMetrics

to track language model stuff like

tokensPerSecond

and

completionTokens

,

TTSMetrics

for text-to-speech details like

audioDuration

, and

STTMetrics

for speech-to-text info such as

audioDuration

. • UsageCollector: This handy tool gathers all your metrics into a

UsageSummary

object, so you can easily track things like

llmCompletionTokens

,

llmPromptTokens

,

ttsCharactersCount

, and

sttAudioDuration

. • Implementation: Just plug the

UsageCollector

into your agent code to start collecting and analyzing those metrics. Here’s a simple example:

from livekit.agents.metrics import UsageCollector

class MyAgent(Agent):
    def __init__(self):
        super().__init__(...)
        self.usage_collector = UsageCollector()

    async def on_agent_metrics(self, metrics):
        self.usage_collector.collect(metrics)

Keep an eye on key metrics like

tokensPerSecond

to check speed and

llmCompletionTokens

for token usage. If you want a hand setting this up or have any questions, just let me know—I’m here to help! Sources: livekit-plugins/livekit-plugins-google/livekit/plugins/google/beta/realtime...

now this is my agent.py help me with the code to improve all the above metrics and token usage, speed etc from dotenv import load_dotenv from livekit import agents from livekit.agents import AgentSession, Agent, RoomInputOptions from livekit.plugins import ( openai, cartesia, deepgram, noise_cancellation, silero, ) from livekit.plugins.turn_detector.multilingual import MultilingualModel load_dotenv() class Assistant(Agent): def __init__(self) -> None: super().__init__(instructions="You are a helpful voice AI assistant. You eagerly assist users with their questions by providing information from your extensive knowledge. Your responses are concise, to the point, and without any complex formatting or punctuation. You are curious, friendly, and have a sense of humor.") async def entrypoint(ctx: agents.JobContext): session = AgentSession( stt=deepgram.STT( model="nova-3", language="multi", detect_language=True, interim_results=True, no_delay=True, sample_rate=16000 ), llm=openai.LLM( model="gpt-4o-mini", temperature=0.8 ), tts=cartesia.TTS( model="sonic-2", voice="4df027cb-2920-4a1f-8c34-f21529d5c3fe", language="en", speed=1.0 ), vad=silero.VAD.load( min_speech_duration=0.05, min_silence_duration=0.3, activation_threshold=0.5, force_cpu=True ), turn_detection=MultilingualModel(), ) await session.start( room=ctx.room, agent=Assistant(), room_input_options=RoomInputOptions( noise_cancellation=noise_cancellation.BVC() ), ) await session.generate_reply( instructions="Greet the user in English with 'Hello, Reuben! What’s the deal today?' and then respond in the detected language of the user's input." ) if name == "__main__": agents.cli.run_app(agents.WorkerOptions(entrypoint_fnc=entrypoint))

To help you boost metrics collection, token usage tracking, and speed in your agent code, here are some handy tips you can try out: • Integrate UsageCollector: The

UsageCollector

class is great for gathering metrics like token counts and audio durations. Here’s a quick example:

from livekit.agents.metrics import UsageCollector

usage_collector = UsageCollector()

@session.on("metrics_collected")
def _on_metrics_collected(ev):
    usage_collector.collect(ev.metrics)

async def log_usage():
    summary = usage_collector.get_summary()
    print(f"LLM Prompt Tokens: {summary.llm_prompt_tokens}")
    print(f"LLM Completion Tokens: {summary.llm_completion_tokens}")

• Keep an Eye on Key Metrics: Tracking

llm_prompt_tokens

,

llm_completion_tokens

,

tts_audio_duration

, and

stt_audio_duration

can give you great insights. • Speed Things Up: Try to position your agent close to LLM, STT, and TTS services to cut down on network delays. Also, watch latency metrics like

ttft

(time-to-first-token) for a smoother experience. • Set Rate Limits: Putting limits on token usage helps you stay on budget and spot usage trends early. • Visualize Your Metrics: Logging or using OpenTelemetry for distributed tracing can really help you keep everything in check. If you’d like, I’m here to help you weave these improvements into your code—just let me know! Sources: examples/voice_agents/basic_agent.py | Voice Assistants | livekit/livekit_composite | DeepWiki | LLMMetrics | LiveKit Agents