What actually ends on December 11

A setting ends. From December 11, the Voice section of ChatGPT's settings offers one conversational mode, Live, instead of three. Anyone still on Advanced or Standard is switched automatically, with no migration step and no new chat required. Old conversations stay where they are.

That is a real change for the minority who kept the old modes. Standard's chain — your speech is transcribed, a language model answers, and the answer is read back — is the pipeline closest to how text ChatGPT works, and users who built long, detailed spoken threads on that chain lose the chain, not just a voice. OpenAI's own community forum already carries complaints along those lines. Treat those posts as what they are: individual users describing their own use, not a measure of how many people object.

The exception: video keeps Advanced

Advanced voice survives December 11 in the one place GPT-Live is not ready. Live video and screen sharing in the ChatGPT app still run on Advanced, and will keep doing so until GPT-Live supports them. OpenAI has not said when that will be.

So the clean version of the story — two old modes retired, one new mode in charge — is not quite the version OpenAI is shipping. What ships is a selector with one option, while Advanced continues behind the video features for as long as they take to move. That may be the sensible engineering order. It also means the December 11 date measures OpenAI's willingness to remove a choice before its replacement supports everything the choice did, not the old model's actual end of life.

The number with no denominator

OpenAI's case for the change rests on a growth claim: voice usage has nearly doubled since GPT-Live launched in July, and virtually all voice users are already on it. The reports reviewed provide no absolute usage totals, measurement definition or legacy-mode user count, so neither statement can be checked from outside — and "virtually all" is OpenAI's characterization, not a quantified share.

On OpenAI's own characterization, the group still using the old modes is a small remainder. OpenAI's stated reason for removing their option is simplification: "This change will allow us to simplify the voice experience and focus on improving GPT-Live." The company also says it is bringing what it learned from the old modes into GPT-Live, including depth and detail of responses, continuity across conversations and better personalization.

What to do before the date

If you use ChatGPT voice and have never changed the setting, the announced selector migration may not require action from you — check which mode your account is on rather than assuming. If you deliberately use Advanced or Standard, the useful test is short: run the conversation you care most about on Live now, while you can still compare the two on the same thread. Compare response detail, interruption behavior, continuity and task completion on conversations representative of your use. After December 11 there is no comparison left to run, and no setting to switch back to. If you use live video or screen sharing, nothing changes yet, and the date to watch is a later one OpenAI has not announced.

Sources

  1. [1] OpenAI, ChatGPT Voice (Help Center) — the Settings > Voice mechanics, the December 11 removal, the default to Live, and the statement that video and screen sharing keep running on Advanced until GPT-Live supports them. ALFW read the article's current text through the reports below, including AI Tool Analysis's screenshot-based reading of the page (updated November 11, 2026); the desk confirmed the article's identity and URL first-hand.Read source
  2. [2] OpenAI Newsroom (@OpenAINewsroom) post on X, November 10, 2026 (status 2120277245713654201) — the announcement that Advanced and Standard voice retire on December 11, the nearly-doubled voice usage claim, and the lessons being carried into GPT-Live; text corroborated across the reports below.Read source
  3. [3] Neowin, "OpenAI is retiring ChatGPT's Advanced and Standard Voice modes, sparking user backlash", November 11, 2026 (full timeline, notification date, history treatment, custom GPT voice until December 11, user objections, OpenAI's acknowledgement)Read source
  4. [4] Interesting Engineering, "OpenAI pulls the plug on two major voice modes as GPT-Live takes over", November 11, 2026 (retirement mechanics, three-model description of Standard, delegation to text models, video and screen-sharing exception, no timeline for the handoff)Read source
  5. [5] AI Tool Analysis, "ChatGPT Voice" (close reading of OpenAI's current voice help page, with screenshots; updated November 11, 2026)Read source
  6. [6] ANI syndicated report via EuropeSays, November 11, 2026 (the direct OpenAI quote: "This change will allow us to simplify the voice experience and focus on improving GPT-Live.")Read source