Takeaway: OpenAI will retire Advanced and Standard voice in ChatGPT on December 11, 2026. A Help Center article updated on November 10 says users were notified that day, can keep using those modes during the transition, and will no longer be able to select them from Settings → Voice after December 11. Anyone still on Advanced or Standard will be switched to Live voice, which is powered by GPT-Live. Live video and screen sharing will keep using Advanced voice until GPT-Live supports those features. The same date is also the scheduled retirement for custom GPTs on most accounts, which is a separate product change.
OpenAI’s reason is consolidation. It introduced GPT-Live in July, says voice usage has nearly doubled since then, and says virtually all people using voice in ChatGPT now use GPT-Live. Retiring the older modes, it writes, will “simplify the voice experience and focus on improving GPT-Live.” If you still prefer Advanced or Standard, you have until December 11 to use them — and you should plan around the features Live does not yet have.
The December 11 cutoff
The retirement article gives a short timeline:
- November 10, 2026: Users are notified about the retirement.
- During the transition: Users can continue using Advanced or Standard voice.
- December 11, 2026: Users will no longer be able to select Advanced or Standard voice from the model picker in Settings → Voice. Users who previously selected those models will be defaulted to Live voice.
Your existing conversations are not being deleted as part of this change. OpenAI says conversations remain in your ChatGPT history and can be continued in text or voice. After December 11, voice in those threads defaults to GPT-Live.
The same Help Center page says voice with custom GPTs will continue using Advanced voice until “their deprecation on December 11, 2026,” and it links to the custom GPT retirement and migration FAQ. That FAQ is a broader change: custom GPTs themselves are scheduled to retire on December 11, 2026, with a later date of February 11, 2027 for FedRAMP workspaces and Enterprise workspaces that have a qualifying, approved deferral. Existing GPT conversations remain accessible after retirement, but GPT pages become inaccessible on the applicable date, and OpenAI is directing creators to migrate to plugins. If you rely on a custom GPT, treat that FAQ as the source of truth — not this voice article.
Live vs Advanced vs Standard
OpenAI’s retirement note and the updated ChatGPT Voice help article describe three different designs. The differences matter if you are deciding whether December 11 is a minor switch or a real workflow change.
| Mode | How OpenAI describes it | What happens on December 11 |
|---|---|---|
| Live | Powered by GPT-Live (the Voice article currently names GPT-Live-1.1 and GPT-Live-1.1 Quip, depending on plan). It is a full-duplex model: it can listen and speak at the same time, offer short backchannels such as “mhmm” or “yeah,” and stay quiet when you need a moment. It can also hand complex questions or agentic tasks to frontier text models in the background. | This becomes the default voice mode. It remains selectable. |
| Advanced | The previous real-time, turn-based voice model. It processed and generated audio natively, but waited for you to stop speaking before it replied. It is also the mode that still supports live video and screen sharing on eligible mobile apps. | Removed from the Settings → Voice picker. Existing users are moved to Live. Video and screen sharing keep using Advanced until GPT-Live has those capabilities. |
| Standard | A three-model pipeline: speech-to-text, a language model, then text-to-speech. OpenAI says this was the original Voice experience and that it produced slower replies and lost information across voice and text. | Removed from the picker. Existing users are moved to Live. |
The Voice help article adds a few Live-specific limits you will not want to discover on December 11. Live can use web search and memory, show visual results through supported widgets, work with text and images in the same chat, and use the plugins and connected apps available to your account. It does not support video or screen sharing. Preset ChatGPT personalities do not currently apply to Live, though you can still ask it to change tone or pace in a single conversation. Precise playback-speed controls are not available.
If you use Voice in Work, or Voice in the desktop app to coordinate agents — including the ChatGPT desktop computer-use features this site covered on November 10 — OpenAI says those experiences are subject to their own plan and workspace settings. The retirement article is about Advanced and Standard in ChatGPT Voice, not a separate announcement that desktop or Work voice is going away.
Video, screen sharing, and other exceptions
This is the sharpest practical exception in the announcement. OpenAI says live video and screen sharing “will continue to be powered by Advanced voice until GPT-Live has this capability.” The Voice help article is consistent with that: Live does not support video or screen sharing, and those features remain available to eligible subscribers in the iOS and Android apps when using Advanced — camera for live video, and the more-options menu for Share Screen.
OpenAI has not given a date for when GPT-Live will pick up video or screen sharing. After December 11, the Settings picker will no longer offer Advanced as a general voice mode. The company is saying the Advanced stack will still sit behind those two features until Live catches up. If video or screen sharing is part of your daily use, confirm the in-product behavior as December 11 approaches rather than assuming the button disappears or that Live suddenly gains a camera.
What OpenAI says it is improving in Live
OpenAI acknowledges that some people are used to Advanced or Standard “style or way of responding,” and that model transitions can be hard. It says it is carrying lessons from those modes into GPT-Live, especially around the depth and detail of responses, continuity across conversations, and personalization. Those are goals, not a changelog of features already shipped. The company also points readers to its Help Center article on navigating model transitions.
GPT-Live itself is not new in this notice. OpenAI introduced GPT-Live in July 2026 as a full-duplex voice model. The November 10 article is the retirement of the older Voice options, not the debut of Live.
Usage limits, plans, and data
Live usage is metered. The Voice help article, which OpenAI also updated around this notice, says Live usage in Chat is measured over a rolling 24-hour period and that ChatGPT will notify you when you reach a limit. The allowances it lists as of this article include limited access to GPT-Live-1.1 Quip on the Free plan; 3 hours of GPT-Live-1.1 Quip on Go; 3 hours of GPT-Live-1.1 on Plus, Business Standard, K-12 Teachers, and legacy Enterprise and Edu plans; 15 hours of GPT-Live-1.1 on the $100 Pro plan and on Business Premium; and unlimited GPT-Live-1.1 on the $200 Pro plan, subject to abuse guardrails. Business plans can buy additional usage at 1.25 credits per minute. Credit-based Enterprise, Edu, and Clinician plans are listed at 1.25 credits per minute, and usage-based Enterprise USD pricing at $0.05 per minute. Those figures can change; check the live Help Center page for your plan.
Voice is available in eligible Business, Enterprise, Edu, and Healthcare workspaces, subject to workspace settings. Businesses in ChatGPT, which OpenAI officially launched in beta on November 10, is a different product: it is about company profiles and Business Agents, not this Voice retirement.
On data, OpenAI says audio clips from Live and Advanced conversations, and video clips from Advanced conversations, are stored with the transcript that appears in chat history and are retained for 30 days. Deleting a chat also deletes the associated clips within 30 days, with the usual legal and safety exceptions. Standard voice transcribes audio and then deletes it after transcription, unless you have chosen to share audio to help train models. OpenAI says it does not use the audio or video clips themselves for training unless you turn on the relevant data-control toggles.
What to do before December 11
- If you still use Advanced or Standard, open Settings → Voice and confirm which mode is selected. After December 11, that picker will no longer offer those two options, and you will be moved to Live.
- If you depend on live video or screen sharing in the mobile apps, keep using Advanced for those sessions during the transition and watch for in-product changes after December 11. OpenAI has not said when GPT-Live will support them.
- If a custom GPT is part of a voice workflow, read the custom GPT retirement FAQ and start migration when it is available for your account. Voice on custom GPTs stays on Advanced only until the December 11 deprecation described in the retirement article.
- Export anything you may need later. OpenAI’s retirement FAQ points to its existing data-export instructions; conversations themselves stay in history, but an export is still the way to keep a copy outside ChatGPT.
- If Live already feels shallower or less personal than Advanced for your use case, say so in feedback and test whether asking Live for a slower, more detailed style helps. OpenAI is promising improvements in those areas; it has not published a before-and-after quality study.
Facts at a glance
| Question | What OpenAI’s Help Center currently says |
|---|---|
| What is retiring? | Advanced and Standard voice in ChatGPT Voice. |
| When? | December 11, 2026. Users were notified on November 10 and can keep using the older modes until then. |
| What replaces them? | Live voice, powered by GPT-Live. People who still have Advanced or Standard selected will be defaulted to Live. |
| Do old chats disappear? | No. Conversations remain in history and can be continued in text or voice. Voice in those threads defaults to GPT-Live after December 11. |
| What about video and screen sharing? | They stay on Advanced voice until GPT-Live supports them. Live does not currently support video or screen sharing. |
| Is this the same as custom GPTs going away? | No, but the scheduled date matches for most accounts. Custom GPTs are covered in a separate FAQ, with a later date for some Enterprise and FedRAMP workspaces. |
| Can I keep Advanced just for its speaking style? | Not after December 11. It will no longer appear in the Settings → Voice picker. |
FAQ
Can I keep using Advanced voice after December 11?
Not as a selectable voice mode. OpenAI says the Advanced and Standard options leave the Settings → Voice picker on December 11, and users who still had them selected are moved to Live. The exception is live video and screen sharing, which OpenAI says will keep using Advanced until GPT-Live can do those jobs.
Is GPT-Live new?
No. OpenAI introduced GPT-Live in July 2026. The November 10 notice is about retiring the older Advanced and Standard modes so ChatGPT Voice can standardize on Live.
What is the difference between Live and Advanced?
Live is full-duplex: it can listen and speak at the same time and can delegate harder work to text models in the background. Advanced was a turn-based real-time audio model that waited for you to finish speaking. Advanced is also the mode that still supports mobile video and screen sharing. Live does not support those features yet.
Will my old voice chats still be there?
Yes. OpenAI says those conversations stay in your ChatGPT history and can be continued in text or voice. After December 11, voice in them defaults to GPT-Live.
Does this affect custom GPTs?
The voice retirement article says voice with custom GPTs will keep using Advanced until their deprecation on December 11, 2026. Separately, OpenAI plans to retire custom GPTs on that date for most accounts and is asking creators to migrate them to plugins. Check the custom GPT FAQ for the timeline that applies to your workspace.
Are Voice usage limits changing on December 11?
The retirement article does not announce new limits. The Voice help article still lists rolling 24-hour Live allowances by plan, including hours on Plus and Pro and credit-based rates on some business and enterprise plans. Use the live Help Center page for the figures that apply to you.
Can I export my voice conversations?
OpenAI points to its existing ChatGPT data-export process. Conversations remain in your history after the voice change, but an export is still the way to keep a copy outside the product.
Will Live copy the way Advanced or Standard sounded?
OpenAI says it is bringing learnings from those modes into GPT-Live, including response depth, continuity, and personalization. It also says some people will find the transition difficult. That is a product goal, not a guarantee that Live will match a voice style you already like.