how to integrate deep infra with voice agent

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

I meant deepseek

Hey there! If you’re looking to integrate DeepSeek with a voice agent in LiveKit, here’s a simple step-by-step guide to help you out: 1. Install the Plugin You Need:

pip install "livekit-agents[openai]~=1.2"

2. Set Up Your Environment Variables: Pop your DeepSeek API key into a

.env

file like this:

DEEPSEEK_API_KEY=<your-deepseek-api-key>

3. Get the DeepSeek LLM Ready: Here’s some code to set up the LLM: `from livekit.plugins import openai deepseek_llm = openai.LLM.with_deepseek( model="deepseek-chat", # Default model temperature=0.7, parallel_tool_calls=True, tool_choice="auto" )`Or, if you want to specify the DeepSeek base URL, try this:

llm=openai.LLM(
    base_url="<https://api.deepseek.com/v1>",
    model="deepseek-chat",
    api_key=os.getenv("DEEPSEEK_API_KEY")
)

4. Plug It Into a VoicePipelineAgent: Use the LLM you just set up in your voice agent pipeline like so:

agent = VoicePipelineAgent(
    vad=ctx.proc.userdata["vad"],
    stt=openai.STT.with_groq(),  # Or another STT provider
    llm=deepseek_llm,
    tts=openai.TTS(),  # Or another TTS provider
    chat_ctx=initial_ctx
)

5. Other Options: Just so you know, DeepSeek models also work with providers like Cerebras, Groq, and others that support OpenAI-compatible APIs. If you hit any bumps or want a hand with any of these steps, just give me a shout—I’m here to help! Sources: chain-of-thought-tts/agent.py | Cerebras and LiveKit | LiveKit Docs | Large language model (LLM) integrations | LiveKit Docs | DeepSeek LLM integration guide | LiveKit Docs | Building an Automated IVR Menu Caller | LiveKit Docs

client = OpenAI( _base_url_="https://api.deepinfra.com/v1/openai", _api_key_=os.getenv("OPENAI_API_KEY"), ) will this be enough

I don't have the answer you're looking for. You could also try asking your question: • in one of the other Slack channels or • to https://deepwiki.com/livekit/livekit_composite which is trained on all LiveKit source code If you find the answer, please post it here to help others!

class Evi(Agent): def __init__(self) -> None: super().__init__(instructions=instructions) self.paused: bool = False async def on_enter(self) -> None: if not self.paused: await self.session.generate_reply( instructions=self.instructions ) async def on_exit(self) -> None: return await super().on_exit() how can I pass metadata to my agent along with instructions. this metadata will be unique to every user