Voice
Conversations, not transcriptions
Echo's voice chat uses OpenAI's realtime API and ElevenLabs voices for a natural back-and-forth. Tap to talk, get a reply in under a second.
Overview
Voice chat with Echo isn't a transcription gimmick. It's a real-time bidirectional conversation powered by OpenAI's gpt-realtime model. Echo hears you, understands tone and pauses, and replies in your chosen voice with sub-second latency. You can interrupt mid-sentence the way you would in any natural conversation. Combined with ElevenLabs voice cloning, you can give your Echo your own voice - or pick from 100+ professional presets. Voice works on every Echo surface: hosted page, embed widget, and (coming soon) WhatsApp voice notes.
What you get
- Realtime back-and-forth - Powered by gpt-realtime. Interrupt, pause, change topics naturally.
- Custom voice - Clone your own voice with ElevenLabs or pick from a wide selection of professional presets.
- Multilingual - Speaks back in the visitor's language without configuration.
- Hands-free - Works on mobile lockscreen with full audio session integration.
How it works
Who it is for
- Hands-free product Q&A - Visitors ask about sizes, ingredients, availability while cooking or driving.
- Personal AI companion - Creators offer voice conversations to fans as a premium tier.
Keep exploring
Related features
Recommended integrations
Also for
Frequently asked questions
How fast is the response?
Typically under 800ms from when you stop talking to when Echo starts replying.
Can I clone my own voice?
Yes, with ElevenLabs voice cloning. Upload a 60-second sample and your Echo speaks as you.
Does it work on mobile?
Yes. Voice chat works in-browser on iOS and Android, and inside the native Echo iOS app.