OpenAI rolls out Advanced Voice Mode for ChatGPT with new voices and emotional awareness

OpenAI has announced the rollout of Advanced Voice Mode for ChatGPT, bringing enhanced voice capabilities to its popular AI chatbot. This new mode, available to ChatGPT Plus and Team subscribers, introduces several improvements, including more natural and real-time conversations, emotional awareness, and five new voices.

Powered by GPT-4o, a new multimodal model that handles text, vision, and audio, Advanced Voice Mode enables faster responses and more nuanced conversations. Users can interrupt the chatbot at any time, and the AI is designed to sense and respond to user emotions.

Personalisation and New Voices

Advanced Voice Mode also supports Custom Instructions and Memory, allowing users to tailor the chatbot's responses to their specific preferences and past interactions. Five new voices – Arbor, Maple, Sol, Spruce, and Vale – have been added, joining the existing four voices (Breeze, Juniper, Cove, and Ember).

Availability and Rollout

OpenAI is gradually rolling out Advanced Voice Mode to all ChatGPT Plus and Team users. Enterprise and Edu subscribers will gain access starting next week. Users will be notified via a pop-up message when the feature is available.

Future Updates and Regional Limitations

Video and screen-sharing capabilities are not included in this initial rollout but are planned for a future update. Advanced Voice Mode is currently unavailable in the EU, UK, Switzerland, Iceland, Norway, and Liechtenstein.

Safety and Testing

OpenAI has conducted extensive safety testing with external red teamers representing diverse languages and geographies. The company has also published a GPT-4o system card detailing its efforts to ensure the safety and user-friendliness of Advanced Voice Mode.

With Advanced Voice Mode, OpenAI aims to create a more human-like and engaging conversational experience with its AI chatbot.