OpenAI set to retire older voice modes in favor of GPT-Live by december 11

OpenAI has announced the discontinuation of its Standard and Advanced Voice Modes in ChatGPT, effective December 11, 2026, as the company consolidates development efforts around the newer GPT-Live model. This move follows the introduction of GPT-Live in July of this year, which features a full-duplex architecture allowing for simultaneous listening and speaking. The transition is driven by GPT-Live's superior capabilities, which have seen usage of ChatGPT's voice functions nearly double since its release. While the older modes are being phased out, certain exceptions remain until the deadline: Advanced Voice Mode will continue to support real-time video and screen sharing, as well as voice interactions with custom GPTs. However, the change has drawn criticism from some users on platforms like X and Reddit. Critics argue that the older Standard Voice Mode offers better contextual continuity and more detailed responses than GPT-Live, which some users report interrupts conversations or provides unhelpful answers. OpenAI acknowledges these concerns and stated it is incorporating feedback to improve GPT-Live's depth and coherence, recalling past instances where user pushback led to the postponement of other feature rollouts.