OpenAI retires ChatGPT's older voice modes on December 11
ChatGPT is about to lose two of its three ways to talk. OpenAI will retire the Standard and Advanced voice modes in December, and a vocal group of users is not happy about it.
Three modes become one
The change lands on December 11 and applies across all ChatGPT subscriptions. After that date, Live will be the only voice mode left. Until then, you pick a mode in Settings, under Voice.
OpenAI says nearly all ChatGPT users already use the newer Live mode. The company is consolidating voice into it so that it can spend its development time on one model instead of three.
Voice in ChatGPT does not disappear. It is a way to talk to the chatbot in real time: you speak, and it answers aloud.
How the modes work differently
Older Standard mode is a relay: speech becomes text, a written answer is produced, and only then is it spoken. Advanced mode handles speech directly, yet speaking and listening still alternate. For that reason it never interrupts.
Live mode can listen and speak at the same time. Interruptions and replies are meant to feel more natural.
Why some users are angry
Some users have taken to Reddit and X to complain. One user wrote, “I'm beyond disappointed because ChatGPT means a lot to me.” Another said they feel like they are paying a monthly subscription to beta-test questionable decisions.
The objections come down to preference. Some people simply like how the older modes work and, above all, how their voices sound. One Reddit poster has used the Vale voice for years and finds it comforting. Picking Vale in Live does not reproduce exactly the same experience.
Others say Live's more natural conversation feels less natural to them. While it thinks, Live pads the silence with “yeah” and “anyway”. Its interruptions can throw a speaker off mid-thought. The tone also shifts more often. Many find that irritating.
Not the first retirement attempt
OpenAI has tried this before. In September 2025 OpenAI set a 30-day deadline for dropping Standard Voice. On September 9 it reversed course. It acknowledged that users valued the older experience and said Standard would stay while it addressed feedback.
This time OpenAI has addressed the objections in its Help center. It said: “We recognize that some people are accustomed to Advanced or Standard voice's style or way of responding, and these model transitions can be challenging.” It points users to a guide on navigating model transitions.
How far the opposition will grow is still unclear.
Source: TechRadar