More natural voice conversations with GPT-Live 1

The new default voice model listens and speaks at the same time, so you can interrupt it.

Date: 2026-10-27

Category: Update

GPT-Live 1 is now the default voice model for voice conversations in AI-School, in the browser and on iOS and Android. The model listens and speaks at the same time, so you can interrupt it to add something or refine your question. This makes the conversation feel more natural.

This is useful, for example, when you’re exploring lesson ideas aloud or practising an explanation. Realise while it’s responding that you want to use the explanation for a different class? You can say so right away, without waiting for the full response.

For more difficult questions and when using tools, GPT-Live 1 uses an underlying text model. You don’t need to select a different model to use the new default voice model.

Does your school use Google for voice? Gemini Live remains available in the browser as an alternative.

Try a voice conversation while preparing your next lesson or practising an explanation.