OpenAI Retires Two ChatGPT Voice Modes on December 11 as GPT-Live Takes Over
OpenAI will retire Advanced and Standard Voice in ChatGPT on December 11, moving remaining users to GPT-Live, the voice model it launched in July. Live video and screen-sharing will keep running on the older system until GPT-Live supports them.
OpenAI said Monday it will retire the Advanced Voice and Standard Voice modes in ChatGPT on December 11, moving remaining users to GPT-Live, the voice model it introduced in July, according to a post from the company's OpenAI Newsroom account on X.
"We're retiring Advanced and Standard voice from ChatGPT Voice on December 11, 2026," the company wrote. "In July, we introduced GPT-Live, our latest voice model that combines natural conversation with frontier intelligence."
OpenAI said voice usage has nearly doubled since GPT-Live launched, and that virtually everyone currently using voice in ChatGPT has already switched to it, making the older modes a shrinking slice of usage.
The two retiring systems work differently from each other and from their replacement. Standard voice relied on three separate models, chained together for speech-to-text, response generation and text-to-speech, a setup that introduced delays between turns. Advanced voice processed audio directly but still required a user to finish speaking before it would generate a response. GPT-Live uses a different architecture that lets the model listen and speak at the same time, according to OpenAI, allowing conversation to flow without the back-and-forth turn-taking of the older modes, Interesting Engineering reported.
Users still on Advanced or Standard voice will be switched automatically to GPT-Live on December 11, and Advanced and Standard will disappear from the model picker under ChatGPT's voice settings. Past voice conversations will remain accessible in chat history. Custom GPTs that currently rely on Advanced voice will also lose access to it that day.
One exception remains: live video and screen-sharing in ChatGPT voice mode will keep running on Advanced voice's infrastructure, because GPT-Live does not yet support those features, OpenAI said. The company said it will maintain the older system for that purpose until GPT-Live gains the same capability.
OpenAI did not say what share of voice users remain on Advanced or Standard voice today, nor did it give a timeline for when GPT-Live might take over live video and screen-sharing. The company also did not disclose whether the change affects voice access or pricing for ChatGPT's free tier.
The retirement is OpenAI's biggest simplification yet of ChatGPT's voice lineup, which has expanded and contracted several times since the company first added voice chat in 2023. It follows OpenAI's push this month to let businesses claim ChatGPT profiles and give some retailers their own AI agent, part of a broader effort to make ChatGPT a more central, everyday tool rather than a chat window people open occasionally.

