Test the VoiceMode API directly from your browser.
Upload an audio sample to create a cloned voice profile.
WAV, MP3, M4A, OGG, FLAC, WebM. Max 25MB. Best results with 15+ seconds of clear speech.
The underlying OpenAI voice profile. Auto-detect assigns one based on your audio.
Generate speech from text using a cloned voice. Returns audio stream.
From /api/voices/clone response. Or select from your voices.
Max 4096 characters.
Real-time voice conversation. Speak into your microphone and hear the AI respond with your cloned voice.
Select from your voices or clone a new one.
Optional. Customize the AI's personality and behavior.
List all your cloned voices.