OpenAI has announced the rollout of an advanced voice mode for ChatGPT. The new feature is now available to users with ChatGPT subscriptions starting at $20 per month. It offers more natural and fluid conversations, although users might encounter a rate limit if they use it extensively in one day.
The advanced voice feature is being gradually released this week. It is currently unavailable in the EU, Iceland, Liechtenstein, Norway, Switzerland, and the U.K. OpenAI first revealed this capability in May, attracting attention due to a voice called Sky, which resembled Scarlett Johansson’s character from the 2013 movie “Her.” Legal issues, however, forced OpenAI to halt the use of this particular voice. Since then, users on the free tier could configure ChatGPT with other voices.
The advanced version now offers quicker response times and pauses to listen if interrupted. Nine voices are available, and users can customize voice chats through the app’s settings. “It’s an increasingly competitive space,” noted Sam Altman, OpenAI’s co-founder and CEO, in a statement.
The competition includes Meta, which has been rolling out its own voice feature on Android devices and planning to introduce celebrity voices accessible through Facebook, Instagram, and WhatsApp. OpenAI, which launched ChatGPT in late 2022, reported over 100 million weekly active users by August 2024. The advanced voice mode is exclusive to subscribers of OpenAI’s Plus, Team, or Enterprise plans.
The Plus tier is the most affordable, at $20 per month. To get started with the new voice mode, ensure you have the latest version of the ChatGPT app on your phone. Open the app, and you will receive a notification once access is granted.
After hitting the continue button, create a new chat by swiping right or tapping the two-line icon in the top left corner and selecting ChatGPT at the top. Tap the sound wave icon next to the “Message” text field and ensure your sound is on. Start speaking when the screen animates in a fluid sky-like blue and white pattern.
Advanced voice subscriptions for ChatGPT
The new feature allows interactions such as bedtime stories, job interview preparations, and language practice, though access is not unlimited. Users reported a “15 minutes left” warning after about half an hour of usage.
OpenAI has not provided additional details on this time limit. As part of the rollout, AVM is getting a revamped design, represented by a blue animated sphere instead of the previous animated black dots. Users will receive a pop-up notification in the ChatGPT app, next to the voice icon, when AVM becomes available to them.
The rollout will extend over the course of the week. OpenAI is introducing several new elements with this update, including Custom Instructions, Memory features, and five new voices named Arbor, Maple, Sol, Spruce, and Vale, bringing the total number of voices to nine. These nature-inspired names align with AVM’s goal of creating a more natural user experience.
While the AVM update does not include the video and screen sharing capabilities introduced during the spring update, which allow GPT-4 to process visual and audible information simultaneously, OpenAI has not provided a timeline for when these multimodal features will be relaunched. OpenAI has made several improvements to AVM based on feedback from its limited alpha test. Enhancements include better accent recognition and smoother, faster conversations.
The Custom Instructions feature allows users to personalize their interactions with ChatGPT, and the Memory feature enables the bot to remember past conversations for future reference. To mitigate potential misuse of the feature, OpenAI has limited the Advanced Voice Mode to preset voices created in collaboration with paid voice actors. These preventive measures ensure the feature can’t be used to impersonate specific individuals or public figures.
Additionally, the company has implemented guardrails to block requests for violent or copyrighted content and included filters to block the generation of music or copyrighted audio. This move has likely been influenced by legal actions from the music industry against AI companies. Many U.S. consumers are willing to pay for smart, reliable voice assistants, according to a PYMNTS Intelligence report.
The report found that 22% of Gen Z consumers are highly willing to pay more than $10 per month for a voice assistant that is as smart and reliable as a real person. Additionally, 54% of consumers prefer using voice technology because it is faster than typing or using a touchscreen.
