OpenAI will end ChatGPT’s standard voice mode and advanced voice mode from December and leave only Live voice mode.

Tech outlet TechRadar reported on Nov. 11 that the measure will apply to all ChatGPT subscription types.

ChatGPT voice mode works by having ChatGPT respond in real time when a user asks a question. OpenAI has opted to consolidate its voice features into a single model, saying most users already use Live voice mode. Some users who are used to the existing options are pushing back strongly.

The change is scheduled for Dec. 11. Until then, users can choose among the three modes in the voice menu in settings. After that, only Live voice mode will be available.

The three modes work differently. Standard voice mode converts a user’s voice into text, generates an answer and then reads it back in voice. Advanced voice mode processes voice directly but alternates between listening and speaking. Live voice mode processes listening and speaking at the same time, making it more natural in how it understands and responds to conversational context.

Some users expressed dissatisfaction with the changed experience rather than performance improvements. Users accustomed to specific voices said that even when selecting the same voice in Live voice mode, it sounds different. One Reddit user said, "It's so disappointing for ChatGPT," and another said, "I pay a subscription fee every month, but it feels like the company tests users every time."

There were also complaints about the conversation style. Live voice mode aims for natural conversation, but some users said unnecessary interjections and interruptions break the flow. They said it inserts phrases such as "yeah" or "anyway" while users are choosing their words, or cuts in mid-sentence and disrupts speech. Some users also said frequent shifts in tone make it feel unnatural.

OpenAI is also aware of the backlash. The company said it will provide guidance documents on the transition process.

OpenAI also announced in September 2025 that it would end standard voice mode within 30 days. But in September that year, it reversed course after acknowledging that users valued the existing experience. OpenAI said at the time it would keep standard voice mode while incorporating user feedback.

The voice feature itself is not disappearing. OpenAI made clear it is choosing to focus development resources on Live voice mode rather than keeping both standard and advanced modes. That leaves the key question of whether user backlash will lead to an actual policy change. With OpenAI having previously withdrawn a plan to end the mode, further adjustments remain possible if demands from users who prefer the existing voice style continue to grow.