OpenAI announced on Tuesday that it is broadening the availability of its Advanced Voice Mode (AVM) feature for ChatGPT, extending it to a larger group of paying customers. Initially, this feature will be accessible to subscribers of ChatGPT’s Plus and Teams plans, with Enterprise and Edu users anticipated to gain access in the following week.

As part of this update, AVM will feature a new design, now represented by a blue animated sphere, replacing the black animated dots that were initially showcased during OpenAI’s technology demonstration in May.

“Users will see a notification in the ChatGPT app, adjacent to the voice icon, when AVM becomes available to them,” OpenAI stated on Tuesday via its X account #OpenAI. “Advanced Voice is being rolled out to all Plus and Team users in the ChatGPT app throughout the week.

“While you’ve been waiting, we’ve introduced Custom Instructions, Memory, five new voices, and enhanced accents. It can also express ‘Sorry I’m late’ in over 50 languages,” the announcement continued.

The update “introduces five new voices—Arbor, Maple, Sol, Spruce, and Vale—bringing the total number of available voices to nine.” These nature-inspired names are intended to embody the natural essence that OpenAI seeks to achieve with AVM.

The previous voices, Breeze, Juniper, Cove, and Ember, will still be available. However, the voice known as Sky, which was highlighted during OpenAI’s spring update, is notably missing from the new selection. Sky was removed following a legal dispute with actress Scarlett Johansson, who argued that the voice closely resembled her own. OpenAI denied any intention to replicate her voice but decided to withdraw Sky after the incident.

One feature that has not been included in this latest AVM rollout is the video and screen-sharing capability that OpenAI first introduced four months ago during its spring update. This feature, which enables ChatGPT to process both visual and auditory inputs simultaneously, currently lacks a release timeline.

During the demonstration, an OpenAI representative showcased how users could pose real-time inquiries to ChatGPT regarding math problems or code visible on their screens. However, there has been no subsequent information regarding the timeline for the availability of this feature.

OpenAI has indicated that several enhancements have been made since the initial limited alpha release of AVM. The voice functionality now includes improved accent recognition, and interactions are said to be more fluid and quicker. While users experienced some technical issues during earlier testing phases, OpenAI asserts that these problems have been largely addressed.

In addition to the voice enhancements, ChatGPT is also broadening its customization options within AVM. The Custom Instructions feature empowers users to modify how ChatGPT interacts with them, while the Memory feature allows ChatGPT to remember previous conversations for future use.

Despite these improvements, AVM remains unavailable in certain regions, including the EU, U.K., Switzerland, Iceland, Norway, and Liechtenstein, as noted in the OpenAI announcement.