Old 2 modes replaced by GPT-Live on Dec. 11

OpenAI on Nov. 10 (Nov. 11 Japan time) announced that "Advanced voice" and "Standard voice," the options in ChatGPT's voice conversation feature "ChatGPT Voice," will be replaced by the voice model "GPT-Live" on Dec. 11. From that day, the 2 modes can no longer be selected under voice in settings, and users who had selected them will be switched to GPT-Live (shown as Live in settings). From Dec. 11, GPT-Live will be the default for voice conversations in ChatGPT. The company's communications account "OpenAI Newsroom" announced it on X, and notices were posted in the help center and release notes.

We're retiring Advanced and Standard voice from ChatGPT Voice on December 11, 2026. ... virtually all people using voice in ChatGPT now use GPT-Live. ... Live video and screen-sharing will continue to be powered by Advanced voice until GPT-Live has this capability.

@OpenAINewsroom · November 11, 2026

In the help center, OpenAI said voice usage has nearly doubled since it introduced GPT-Live in July and that nearly all voice users now use GPT-Live. Both are the company's own figures, and it did not give the baseline date or the actual number of users.

GPT-Live is a voice model that can speak while listening to the user at the same time, and OpenAI also released it on the developer API in September. According to TheNextWeb, questions that require search or complex reasoning are handed off to a separate model in the background, and the model answers without stopping the conversation. According to reports, Advanced voice handles audio directly but waits for the user to finish speaking before responding. Standard voice chains together 3 models for transcribing speech, generating responses and synthesizing speech.

Live video and screen sharing stay on Advanced voice for now

GPT-Live does not support live video, in which users talk while showing camera footage, or screen sharing. These 2 features will continue on Advanced voice until GPT-Live supports them. Eligible subscribers select Advanced voice in the ChatGPT apps for iOS and Android to use them. OpenAI did not say when GPT-Live will support them.

According to Plain AI Daily, GPT-Live gained the ability to use ChatGPT plugins and connected apps by voice from Sept. 23. However, it does not yet support finding or adding files saved in ChatGPT's "Library."

Old modes usable during transition, and conversations stay in history

OpenAI began notifying users on Nov. 10. During the transition period until Dec. 11, Advanced voice and Standard voice can still be used.

Conversations held in the old modes remain in history and can be continued by text or voice. If continued by voice after Dec. 11, GPT-Live will respond.

Voice conversations with custom GPTs run on Advanced voice. OpenAI also plans to end custom GPTs themselves on Dec. 11, and voice for custom GPTs will use Advanced voice until then. According to reports, custom GPTs will end on Feb. 11, 2027, in the FedRAMP environment for the U.S. government and for some Enterprise customers granted an extension.

The path to consolidating ChatGPT voice into GPT-Live (local dates)
  1. Plan to replace Standard voice announced, later reversedReport
  2. GPT-Live released in ChatGPTOfficial
  3. GPT-Live 1 released on the developer APIOfficial
  4. GPT-Live-1.1 rolled out to all ChatGPT usersReport
  5. Replacement of the old 2 modes with GPT-Live announcedOfficial
  6. Old 2 modes removed as options; custom GPTs endOfficial

Available models and time differ by plan

The help article on the replacement is written for ChatGPT users in general, without limiting it to specific plans. It notes that the available voice modes vary by plan, workspace settings, region, app version and parental controls.

According to the ChatGPT Voice help article, the free tier can use the smaller "GPT-Live-1.1 Quip" in limited amounts. "Go," at $8 a month, gets 3 hours of Quip, and "Plus," at $20 a month, gets 3 hours of "GPT-Live-1.1." "Pro" at $100 a month gets 15 hours, and Pro at $200 a month is unlimited.

Business, Enterprise, Edu and Healthcare workspaces can also use Voice, depending on administrator settings. The help article's table also has rows for organizational plans, and according to Plain AI Daily, which reproduced the help article, Business Standard gets 3 hours and Business Premium 15 hours, with usage beyond that consuming 1.25 credits per minute. Enterprise with pay-as-you-go billing in dollars costs $0.05 per minute, the same level as GPT-Live 1's audio pricing on the developer API.

When GPT-Live was released in July, "GPT-Live-1" was assigned to Go, Plus and Pro, and the smaller "GPT-Live-1 mini" to the free tier. According to Data Studios, it switched to the improved "GPT-Live-1.1" on Oct. 15 and added 13 voices.

Plan to replace Standard voice in 2025 also reversed after backlash

OpenAI also announced a plan to end Standard voice in September 2025 but reversed it after user backlash, TechRadar and others reported.

This time as well, some Plus and Pro users are asking for Standard voice to be kept, Data Studios reported.

OpenAI has been expanding voice features around GPT-Live, including voice support for ChatGPT's household finance feature "Finances" and the general availability of outbound phone calls for developers. With the replacement of the old 2 modes, the foundation of ChatGPT's voice will be concentrated in GPT-Live. When GPT-Live supports live video and screen sharing will determine when ChatGPT's voice is fully unified on GPT-Live.