Have natural, real-time voice conversations with AI. Voice Mode transforms your experience into hands-free, spoken dialogue — just like talking to a person.
Voice Mode enables natural voice conversations with your AI assistant. Simply speak your questions and hear AI responses in real-time. The conversation flows naturally — you can interrupt, ask follow-up questions, and have back-and-forth dialogue.
Voice Mode requires microphone access. Grant permission when prompted in your browser or app settings to start voice conversations.
Voice Mode is designed for spoken conversation: brainstorming, quick questions, practice, dictation, and hands-free help. For image, video, music, or other creation tasks, use chat or the dedicated AI tools.
Getting started with Voice Mode is simple:
Voice Mode uses advanced conversational AI technology to create a seamless voice experience: Key technical features:
The animated orb shows the current conversation state: A spinning indicator appears when the AI is "thinking" — processing your input before responding.
Listening — Ready for you to speak
Speaking — AI is responding
Connecting or Reconnecting
Error — Something went wrong
Control your voice session with these options: On the floating control bar:
Toggle mute/unmute microphone
End the voice session
Start voice mode (from chat)
Choose from multiple AI voices to personalize your experience:
Soft, friendly
Calm, professional
Deep, warm
Energetic, youthful
British, authoritative
Your voice preference is saved automatically. The next time you start a voice session, it will use your previously selected voice.
Voice Mode excels in these scenarios:
For best results, use Voice Mode in a quiet environment. Background noise can affect speech recognition accuracy.
Voice Mode supports natural interruption, called "barge-in". If the AI is speaking and you start talking, it will: This makes conversations feel natural — you don't have to wait for the AI to finish if you want to redirect, clarify, or move on.
All voice conversations are automatically saved to your chat history: When you end a voice session, you're automatically taken to the chat view where you can see the full transcript and continue the conversation via text if you prefer.
Voice Mode uses a flat rate billing model based on session duration: Credits are deducted based on active conversation time. If the connection drops or you're temporarily disconnected, billing pauses until you reconnect. Credit thresholds:
Your remaining credits are displayed during the voice session. You'll receive a warning when credits are running low, and the session will end gracefully if credits are depleted.
Common issues and solutions:
Check browser permissions, ensure no other app is using the microphone, and try refreshing the page.
Make sure you're unmuted, speak clearly, and check your internet connection. The status dot should be purple (listening).
Auto-reconnection handles brief drops. If disconnections persist, check your network stability or try a different location.
Voice Mode requires at least 30 credits to start. Purchase more credits or upgrade to Plus for monthly credit allocation.
Voice Mode works best in Chrome, Edge, and Safari. Some older browsers may have limited microphone support.
Voice Mode requires microphone access. Grant permission when prompted in your browser or app settings to start voice conversations.
Voice Mode is designed for spoken conversation: brainstorming, quick questions, practice, dictation, and hands-free help. For image, video, music, or other creation tasks, use chat or the dedicated AI tools.
Your voice preference is saved automatically. The next time you start a voice session, it will use your previously selected voice.
For best results, use Voice Mode in a quiet environment. Background noise can affect speech recognition accuracy.
Your remaining credits are displayed during the voice session. You'll receive a warning when credits are running low, and the session will end gracefully if credits are depleted.
Voice Mode works best in Chrome, Edge, and Safari. Some older browsers may have limited microphone support.