OpenAI will retire two of ChatGPT's three voice modes next month. Advanced Voice and Standard Voice will disappear from the app's Settings on 11 December 2026, and anyone still using them will be switched automatically to GPT-Live, the company's newer voice model. OpenAI began notifying users about the change on 10 November 2026. ## Why OpenAI is making the switch GPT-Live, launched in July 2026, uses a full-duplex design, which means it can listen and speak at the same time. That allows users to interrupt it mid-sentence, while it can also acknowledge a speaker briefly or stay silent while they think. The older modes were turn-based: they waited for the user to finish speaking before responding. Standard Voice also relied on three separate models for speech-to-text, language and text-to-speech, which added delays. OpenAI says voice use in ChatGPT has nearly doubled since GPT-Live arrived, and most people who talk to ChatGPT now use it. The company says consolidating on one voice model will let it focus development on GPT-Live, including better response depth, continuity across conversations and personalisation. ## What stays the same, and what users lose Until 11 December, both older modes remain available, and past voice conversations will stay in ChatGPT history and can be continued by text or with GPT-Live. Live video and screen sharing will keep running on Advanced Voice for now, because GPT-Live cannot do either yet, and OpenAI has given no date for adding them. Voice chats with custom GPTs will also continue on Advanced Voice until the deadline. The decision has drawn complaints from some users, who say GPT-Live interrupts too often and gives shorter, less detailed answers than Standard Voice. [Neowin](https://www.neowin.net/news/openai-is-retiring-chatgpts-advanced-and-standard-voice-modes-sparking-user-backlash/) and [Interesting Engineering](https://interestingengineering.com/ai-robotics/chatgpt-drops-two-voice-modes-december-2026)