OpenAI rolls out Advanced Voice Mode with more voices and a new look


OpenAI announced it is rolling out Advanced Voice Mode (AVM) to an expanded set of ChatGPT’s paying customers on Tuesday. The audio feature, which makes ChatGPT more natural to speak with, will initially roll out to customers in ChatGPT’s Plus and Teams tiers. Enterprise and Edu customers will start receiving access next week.

As part of the rollout, AVM is getting a revamped design. The feature is now represented by a blue animated sphere, instead of the animated black dots that OpenAI presented during its showcase of the technology in May.

Users will receive a pop-up in the ChatGPT app, next to the voice icon, when AVM has been made available to them.

Advanced Voice is rolling out to all Plus and Team users in the ChatGPT app over the course of the week.

While you’ve been patiently waiting, we’ve added Custom Instructions, Memory, five new voices, and improved accents.

It can also say “Sorry I’m late” in over 50 languages. pic.twitter.com/APOqqhXtDg

— OpenAI (@OpenAI) September 24, 2024

ChatGPT is also getting five new voices that users can try out: Arbor, Maple, Sol, Spruce, and Vale. This brings ChatGPT’s total number of voices to nine (almost as many as Google’s Gemini Live), alongside Breeze, Juniper, Cove, and Ember. You might notice all of these names are inspired by nature, which could be because the whole point of AVM is to make using ChatGPT feel more natural.

One voice missing from this lineup is Sky, the voice OpenAI showcased during its spring update, which led to a legal threat from Scarlett Johansson. The actress, who played an AI system in the feature film “Her,” claimed that Sky’s voice sounded a little too similar to her own. OpenAI promptly took Sky’s voice down, saying it never intended to resemble Johansson’s voice, despite several staff members making references to the movie in tweets at the time.

Another feature missing from this rollout: ChatGPT’s video and screen sharing that OpenAI debuted during its spring update four months ago. That feature is supposed to let GPT-4o simultaneously process visual and audible information. During the demo, an OpenAI staff member showed how you could ask ChatGPT real-time questions about math on a piece of paper in front of you, or code on your computer screen. At this time, OpenAI is not offering a timeline for when it will launch these multimodal capabilities.

That said, OpenAI says it has made some improvements since releasing its limited alpha test of AVM. ChatGPT’s voice feature is allegedly better at understanding accents now, and the company says its conversations are smoother and faster as well. During our tests with AVM, we found that glitches were not uncommon, but the company claims that’s now improved.

OpenAI is also expanding some of ChatGPT’s customization features to AVM: Custom Instructions, which allows users to personalize how ChatGPT responds to them, and Memory, which allows ChatGPT to remember conversations to reference later on.

An OpenAI spokesperson says AVM is not yet available in several regions, including the EU, the U.K., Switzerland, Iceland, Norway, and Liechtenstein.



Leave a Reply

Your email address will not be published. Required fields are marked *