OpenAI announced on Twitter that it will begin rolling out its Advanced Voice feature, along with five new voices for its conversational AI, to Plus and Teams subscribers throughout this week. Enterprise and Edu subscribers will gain access starting next week.
Advanced Voice will be made available to all Plus and Team users in the ChatGPT app over the coming days. While you’ve been waiting, we’ve enhanced the experience with Custom Instructions, Memory, five new voices, and improved accents. The AI can now say “Sorry I’m late” in over 50 languages.
Powered by the advanced GPT-4o model, Advanced Voice allows users to interact with the chatbot verbally, mimicking a natural conversation. It was previewed during OpenAI’s Spring Update event and initially released to select ChatGPT Plus subscribers for beta testing in July. Now, every paying subscriber can experience this feature firsthand.
OpenAI also introduced five new voices for the chatbot: Arbor, Maple, Sol, Spruce, and Vale. These will be available in both Standard and Advanced Voice modes, joining the existing voices — Breeze, Juniper, Cove, and Ember. It’s worth noting that while video and screen sharing features are not currently supported in Advanced Voice, they are expected to be available in the future.
Additionally, OpenAI is integrating two new tools to align Advanced Voice capabilities with the overall chatbot experience: memory and custom instructions. Initially, Advanced Voice could only reference the ongoing chat. With the new memory feature, the AI will now remember details from past conversations, minimizing the need for users to reiterate information. Custom instructions will allow users to set parameters for how the AI responds. For instance, you might specify that code-related responses be delivered in Python.
Plus and Teams subscribers will receive an in-app notification as soon as the feature is activated on their accounts. However, Advanced Voice is currently unavailable in the EU, the U.K., Switzerland, Iceland, Norway, and Liechtenstein. ChatGPT isn't alone in its ability to hold direct conversations with users; this announcement arrives just under two weeks after Google launched its Gemini Live feature for all users, including free-tier subscribers.