BringYourOwnModelQuickstart

Bring-Your-Own-Model Quickstart

For common setup instructions, troubleshooting, and detailed information, see the C# Samples README

This sample demonstrates how to build a real-time voice assistant using direct VoiceLive model integration with bring-your-own-model. It provides a straightforward approach without agent overhead, ideal for scenarios where you want full control over model selection and instructions but with your own model hosted in Foundry.

What Makes This Sample Unique

This sample showcases:

• Bring-Your-Own-Model Integration: Connects direct to a self hosted model
• Proactive Greeting: Agent initiates the conversation with a welcome message
• Custom Instructions: Define your own system instructions for the AI
• Flexible Authentication: Supports both API key and Azure credential authentication

Prerequisites

• AI Foundry resource
• API key or Azure CLI for authentication
• See C# Samples README for common prerequisites

Quick Start

1. Update appsettings.json:

{
  "VoiceLive": {
    "ApiKey": "your-voicelive-api-key",
    "Endpoint": "https://your-endpoint.services.ai.azure.com/",
    "Model": "your-model-name",
    "Byom": "byom-azure-openai-chat-completion", // For multimodal models use "byom-azure-openai-realtime"
    "Voice": "en-US-AvaNeural"
  }
}

2. Run the sample:

   dotnet build
   dotnet run

Command Line Options

# Run with API key (from appsettings.json)
dotnet run

# Run with Azure authentication
dotnet run --use-token-credential

# Run with custom model and instructions
dotnet run --model "your-model-name" --byom "byom-azure-openai-chat-completion"" --instructions "You are a helpful assistant"

# Run with custom voice and verbose logging
dotnet run --voice "en-US-JennyNeural" --verbose

Available Options

• --api-key: Azure VoiceLive API key
• --endpoint: Azure VoiceLive endpoint URL
• --model: VoiceLive model to use (default: "gpt-realtime")
• --byom: BYOM integration mode (default: "byom-azure-openai-chat-completion"; use "byom-azure-openai-realtime" for multimodal models)
• --voice: Voice for the assistant (default: "en-US-AvaNeural")
• --instructions: Custom system instructions for the AI
• --use-token-credential: Use Azure authentication instead of API key
• --verbose: Enable detailed logging

Available Models

• gpt-realtime - Latest GPT-realtime model (recommended)
• gpt-4.1 - GPT-4.1 LLM model
• See documentation for all available models

See C# Samples README for available voices, troubleshooting, and additional resources.