Agent update
ChatGPT Voice powered by GPT-Live
ChatGPT Voice is now powered by GPT-Live, launched on 8 Jul 2026, so you can talk over each other more naturally, pause without being cut off, and have harder questions handed to a later model in the background while the conversation keeps going.
How to use ChatGPT Voice with GPT-Live for small business
A tradesperson on the way to a quote could talk through the job out loud and ask Voice to pull together a rough scope, then keep chatting while a harder pricing question is researched in the background. A clinic receptionist between patients could dictate a follow-up note hands-free, interrupt to add a detail, and expect Voice to wait if they pause mid-sentence. A business owner commuting could practise a client call, ask it to slow down, or glance at a visual card without leaving the conversation.
GPT-Live is rolling out globally on iOS, Android and ChatGPT.com. GPT-Live-1 is the default for Go, Plus and Pro users, and GPT-Live-1 mini is the default for Free users. At launch it does not support voice with video or screen sharing. Some languages may have a non-native accent or gaps in fluency. It uses predefined voices and is not meant to imitate a real person's voice.
Most useful for: Customer Service Representative – Answer customer calls about bookings, orders and accounts, Sales Representative – Follow up leads by phone, Front Desk Admin – Answer the phone during busy periods
Frequently asked questions
Which Voice model do Free and paid plans get?
Free users get GPT-Live-1 mini as the default Voice model. Go, Plus and Pro users get GPT-Live-1. The announcement points to the Help Centre for more availability detail, and it does not state a separate price for Voice.
Can I use it with video or screen sharing, or clone my own voice?
Not at launch. GPT-Live does not support voice with video or screen sharing in ChatGPT, though OpenAI says it is working on those. It uses a set of predefined, remastered voices, with safeguards to stop it imitating a real person's voice. Legacy Standard and Advanced Voice Mode remain available where those older features still exist.
How is GPT-Live different from Advanced Voice Mode?
GPT-Live uses a full-duplex design, so it can listen and speak at the same time. You can talk over each other, pause without being cut off, and have harder questions handed to a later model in the background while the conversation keeps going. At launch it uses GPT-5.5 in the background, and OpenAI says it will keep that current as new models are released.