Voice DevTools

This UI provides a debug console for real-time AI voice interactions. It works with multiple realtime models (View supported models). Features include:

1. Cost Tracking: Know how much you’ve spent per voice interaction
2. Model Support: Supports open-source (MiniCPM-o) and closed-source S2S models (adding more soon!)
3. Metrics Support: Track errors, response timelines, and performance
4. Voice and Chat UI

Inspired by openai-realtime-console and openai-realtime-agents.

Quick Start

1. Get your API keys:

   • Outspeed API key
   • OpenAI API key

2. Set up environment:

   cp .env.example .env
   # Add your API keys to .env:
   # OPENAI_API_KEY="<your-openai-key>"
   # OUTSPEED_API_KEY="<your-outspeed-key>"
   

3. Install and run:

   npm install
   npm run dev
   

Visit http://localhost:3000 to access the console.

Usage

To modify agent prompt and tools, modify agent-config.js.

To modify the model parameters like (voice, version, etc.), edit model-config.js

Agents

There are two voice agent examples already present in agent-config.js:

1. Dental Agent: Answers callers’ questions about working hours of a dental clinic
2. Message Agent: Takes callers’ messages for a person

You can modify the export in the file to activate an agent:

// agent-config.js
export { dental_agent as agent }

// for message_agent
export { message_agent as agent }

Play around with the prompts, or add your own voice agent to give it a spin.

Supported Models

• MiniCPM-o (hosted by Outspeed)
• OpenAI Realtime models
• Moshi (Coming Soon)
• Gemini Multimodal Live (Coming Soon)

License

MIT