OpenAI will switch off ChatGPT’s Standard and Advanced Voice modes on 11 December 2026. Some paying subscribers are asking the company to think again.
What changes on 11 December
Both modes will vanish from the voice settings for all subscription tiers, leaving GPT-Live as the only voice option. OpenAI began notifying users on 10 November. Live video and screen sharing will keep using Advanced Voice until GPT-Live supports them. OpenAI says the move simplifies the experience and lets it focus on improving GPT-Live.
Why OpenAI says it is safe
The company says voice usage has nearly doubled since GPT-Live launched in July, and that most voice users have already moved over. It also acknowledges that some people may prefer the response style of the older modes.
What users are saying
Complaints are spread across Reddit, X and OpenAI’s community forum. Several themes recur.
Many users who prefer Standard Voice describe it as calm, steady and focused. They say GPT-Live feels too human-like and too chatty. One Pro subscriber said GPT-Live is somewhat better than the original Advanced Voice but still too theatrical for them.
Others say familiar voices sound different in GPT-Live even when they select the same name. Some say interjections and interruptions break their flow. Subscribers on paid plans also say they feel treated as testers for decisions they did not ask for.
What they are asking for
The requests are modest. Users want Standard Voice kept as an optional mode, or at least as a legacy read-aloud voice. Others want a plainer speaking style inside GPT-Live that favours speed and clarity over simulated conversation.
A familiar pattern
OpenAI announced plans to retire Standard Voice in September 2025, and the mode has remained available since. That history may explain why some users now expect another reprieve. Nothing in the current announcements suggests one.
What it means
The company has data on its side, since most people already use GPT-Live. Its critics have a different argument: the real question is whether a new voice model that suits most users should replace every older option. For users who rely on voice for accessibility or daily routine, a fixed deadline is hard to accept. Whether OpenAI adds a calmer style before 11 December is the thing to watch.
