← All templatestemplates / general-purpose-conversational-voice-agent-with-nvidia-nim

Pipecatupdated Jan 15, 2025 · other · other

General Purpose Conversational Voice Agent with NVIDIA NIM

A blueprint demonstrating a conversational AI voice agent built with Pipecat framework using NVIDIA NIM infrastructure. The agent uses Meta's Llama 3.3 70B Instruct model for natural language understanding and NVIDIA Riva for speech-to-text and text-to-speech. This example showcases vendor-neutral voice agent architecture with customizable system prompts for various conversational applications.

Try the demo View source Fork template

The numbers

latency—

cost / min—

frameworkPipecat

The stack

telephonyWeb Only

speech-to-textDeepgram Nova-3

llmLlama 3.3 70B

text-to-speechCartesia Sonic Turbo

System prompt

No prompt published.

Config

config.json

{
  "vad": "silero",
  "transport": "daily",
  "environment": "jupyter_notebook",
  "python_version": "3.13",
  "deployment_platform": "brev"
}

Voice AI recipes, picks, and analysis.

Get the useful new templates plus the occasional teardown of what’s working in production voice AI.