This feature uses OpenAI's Realtime API for continuous voice streaming. Speak naturally with the AI with minimal latency.
Session Status:Inactive
Press the microphone button and start speaking. The AI will respond naturally as your conversation flows.
Click to speak
Activity Log
How to Use This Page
Start a Conversation: Click the "Start Session" button, the first time you will have to approve microphone access in browser and then "Click to Speak" to begin a voice chat with the AI.
Speaking to the AI: When the microphone icon pulses red, your voice is being detected and sent to the AI. Pause once the AI is done talking and it will wait till it hears your voice again.
Microphone States:
Gray - Microphone inactive or session not started
Yellow - Waiting for your voice
Pulsing Red - Your voice is being detected
Toggle Microphone: Use the microphone button to temporarily disable/enable your microphone.
Change Character: Select different AI characters from the dropdown menu. Make sure to stop the session select the character and start session.
Debug Panel: Access technical details by clicking "Show Debug Panel".
End Conversation: Click red mic and stop session to end the session.
Note: This is a real-time voice conversation - the AI responds as you speak and you can interrupt the AI at any time. If 30mins of inactivity session will end. Setting gpt-4o-realtime-preview-2024-12-17 is recommended for better performance and following the character prompt