🔍 Read the full analysis: December 11 ChatGPT Update: What Happens To Voice Modes? on ThorstenMeyerAI.com
Get hardware and tech essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
OpenAI says ChatGPT’s Advanced and Standard voice modes will be removed on December 11, 2026, and users will be switched to GPT-Live. Existing conversation histories will remain, while live video and screen sharing will continue using Advanced voice until GPT-Live supports them. The company says nearly all voice users have already moved to GPT-Live, but some users report that its conversational style and voices do not meet their needs.
OpenAI will remove ChatGPT’s Advanced and Standard voice modes on December 11, 2026, leaving GPT-Live as the voice option for users in ChatGPT. The company announced the change on November 10 and says nearly all ChatGPT voice users already use GPT-Live; users who selected either older mode will be switched to it when the retirement takes effect.
OpenAI’s November 10 announcement gives users 31 days’ notice. According to the company’s help-center guidance, as reported by Interesting Engineering and FourWeekMBA, Advanced and Standard remain available during the notice period. On December 11, both options will be removed from Settings > Voice, and people who had selected either one will move to Live voice, powered by GPT-Live. OpenAI says existing conversations will remain in chat history. Users can continue them by text or voice, with voice continuations handled by GPT-Live.
There are stated exceptions. Live video and screen sharing will continue to use Advanced voice until GPT-Live gains those capabilities; OpenAI has not announced a date for that change. The help-center wording also says voice with custom GPTs will continue using Advanced voice until those GPTs are deprecated on December 11. OpenAI’s broader plan to retire custom GPTs across ChatGPT plans is a separate change, according to release notes cited by MIXED.
OpenAI’s stated rationale is that voice use has “nearly doubled” since GPT-Live launched in July and that “virtually all” people using voice in ChatGPT now choose GPT-Live. Those are the company’s descriptions; the cited announcement does not provide the underlying usage figures or a measurement period beyond the comparison with GPT-Live’s arrival.
December 11 voice update What changes for users?
OpenAI plans to remove Advanced and Standard voice from ChatGPT and move users to GPT-Live. Conversation histories will remain available.
What users can expect
The change applies to voice modes inside ChatGPT.
Chats stay available
Existing conversations remain in chat history and can continue by text or GPT-Live voice.
Some features stay on Advanced
Live video and screen sharing will keep using Advanced voice until GPT-Live supports them.
Three modes, one general option
The systems handle speech differently.
| Mode | How it works | After December 11 |
|---|---|---|
| Standard | Speech-to-text, language model, then text-to-speech | Removed |
| Advanced | Processes speech directly; waits for the user to finish | Removed for general voice |
| GPT-Live | Designed to listen and speak at the same time | General voice option |
What the change means
The switch is automatic for users who selected either older mode.
- Spoken continuations will use GPT-Live.
- Some users report concerns about interruptions, voices, and accessibility needs.
- Try GPT-Live before the change if your workflow depends on voice.
- The announcement concerns ChatGPT settings.
- It does not announce an API shutdown.
- Check documentation for the specific API models you use.
Still unclear
OpenAI has not provided a date for several changes.
- When GPT-Live will support live video and screen sharing.
- When older-mode features will reach GPT-Live.
- Whether settings will address reported concerns about interruptions, voices, or accessibility.
How the Switch Affects Voice Users
The change makes GPT-Live the sole general voice mode in ChatGPT, rather than letting users choose among systems that handle speech differently. For people who prefer Advanced or Standard, the switch is not optional after December 11. Their saved conversations remain available, but any spoken continuation will use the new mode.
That matters because voice is not only a convenience for some users. Participants in OpenAI’s community forum have said they rely on voice for accessibility reasons, while other users describe preferring the older voices or interaction style. TechRadar reported complaints that GPT-Live interrupts, fills pauses with phrases such as “yeah” and “anyway,” or sounds unlike voices users had grown accustomed to. These are user reports, not an independently established finding about how the model performs for all users.
For developers and organizations, the distinction between ChatGPT and the API also matters. The announced retirement concerns voice modes inside ChatGPT; it is not, on the available information, an announcement that GPT-Live or all OpenAI voice APIs are being discontinued. Businesses using those APIs should check the separate model documentation and deprecation notices rather than treating this ChatGPT settings change as an API shutdown.
Products Worth Considering
Logitech MX Mechanical Wireless Illuminated Keyboard, Clicky Switches
Ledger Flex + Folio Crypto Hardware Wallet
Apple MacBook Pro 16.2-inch with M5 Pro, 18-core CPU, 20-core GPU, 48GB Memory, and 1TB SSD (Space Black)
As an affiliate, we earn on qualifying purchases.
From Three Voice Systems to GPT-Live
The three modes differ in how they process conversation. Standard voice used a sequence of speech-to-text, language-model, and text-to-speech steps, according to help-center wording quoted by FourWeekMBA. Advanced voice processed speech directly but waited for the user to finish before responding. GPT-Live is designed as a full-duplex system, meaning it can listen and speak at the same time.
OpenAI introduced GPT-Live on July 8 as a new generation of voice models for natural human-AI interaction and said it was powering ChatGPT Voice. Its developer documentation describes the API model GPT-Live-1 as a real-time, full-duplex voice model that can listen and speak while delegating reasoning or tool use to a backend agent. OpenAI has said it plans to bring features from the older modes to GPT-Live and improve response depth, continuity and personalization, according to reports by Android Authority and Interesting Engineering.
The announcement does not mean every voice-related function changes on the same date. Video and screen sharing remain on Advanced voice until GPT-Live supports them. Separately, OpenAI’s API documentation lists older audio and voice models with shutdown dates from January 6 through February 26, 2027, according to the supplied source material. Those dates concern API models, not the December 11 ChatGPT mode retirement.
““I’m beyond disappointed because ChatGPT means a lot to me.””
— A ChatGPT user quoted in TechRadar
Open Questions About GPT-Live
OpenAI has not said when GPT-Live will support live video or screen sharing, or provided a detailed schedule for bringing older-mode features into the new system. The company’s claim that virtually all voice users have moved to GPT-Live is not accompanied in the cited announcement by raw usage totals, a breakdown by mode, or a measurement method.
The scale of opposition is also uncertain. TechRadar described critics as a small but vocal group, while an OpenAI community-forum thread titled “Keep Standard Voice Mode!” had 100 posts when the source article checked on November 12. That count is a snapshot of one discussion, not a measure of all affected users. The cited reports also do not establish whether OpenAI will add settings to address complaints about interruptions, filler phrases, voice choices or accessibility needs.
Finally, the available material does not indicate that the ChatGPT retirement directly changes API access. OpenAI’s API model shutdown dates are separate notices, and developers should verify the status of the specific models they use. The help-center details cited here were reported by outside outlets; the source material says those pages were not machine-readable when checked.
The December 11 Voice Cutover
Until December 11, 2026, users can continue using Advanced or Standard voice, according to the reported help-center instructions. On that date, the two settings are scheduled to disappear and affected users will move to GPT-Live. No manual transfer of chat history is expected; prior conversations remain available, and users can continue them by text or with GPT-Live voice.
Users who depend on voice for accessibility or have workflows built around a particular mode may want to test GPT-Live before the deadline and check whether its interaction works for them. Organizations using voice features should distinguish ChatGPT settings from API products and review OpenAI’s documentation for any model-specific deprecation dates. OpenAI has not announced a date for moving video and screen sharing off Advanced voice.
Key Questions
Which ChatGPT voice modes are being retired?
Advanced and Standard voice are scheduled to leave ChatGPT on December 11, 2026. GPT-Live will remain as the general voice mode.
Will my old voice conversations be deleted?
No deletion is described in the reported help-center guidance. Existing conversations will remain in chat history and can be continued by text or voice; spoken continuations will use GPT-Live.
Will video and screen sharing stop working on December 11?
OpenAI’s help-center wording, as quoted by FourWeekMBA, says live video and screen sharing will continue to use Advanced voice until GPT-Live supports those capabilities. The company has not given a date for that handover.
Does this announcement retire OpenAI’s voice API?
The announced change applies to voice modes in ChatGPT, not a stated blanket retirement of OpenAI’s voice API. Some older API audio and voice models have separate shutdown dates listed between January 6 and February 26, 2027; developers should check the notice for each model they use.
Source: ThorstenMeyerAI.com
