OpenAI Upgrades ChatGPT's Voice To Speak In Different Character Voices

OpenAI is updating ChatGPT’s voice capabilities allowing users to interact with the chatbot using a variety of AI-generated voices and vocal styles.

ChatGPT currently has four preset voices, down from five, since it had to remove “Sky” following a legal complaint from Scarlett Johansson.

The new feature lets ChatGPT users instruct the AI chatbot to respond using any voice in real time.

In a glimpse of what’s to come, OpenAI published a demo where a human asked the chatbot to create voices for characters in a story. ChatGPT responded by generating distinct vocal styles, such as a booming, gruff voice for a lion and a high-pitched, meek squeak for a mouse.

Vocal outputs can also be refined in real time, for example, by requesting that a particular voice sound deeper or higher-pitched. The outputs are generated at speed, with users able to interrupt the bot during a response to request changes.

Users will also be able to continue their ChatGPT conversations when exiting the app.

OpenAI said new voice functionalities will be rolled out “in the coming weeks” but offered few additional details. 

Announced at the company’s recent Spring Update event, new voice and vision capabilities will be made available to all ChatGPT users with premium subscribers given priority access.

Related:Google, OpenAI New AI Agents Shift Focus to Consumers

The voice options extend to the new desktop version too, with earlier demos showing users interacting with the bot while using spreadsheet applications.

OpenAI’s latest voice demo comes as the Microsoft-backed company looks to make its flagship chatbot more interactive and accessible.

Its new GPT-4o model is the driving force behind the new voice and visual features, powering multimodal functions on the mobile app and the new desktop version.

The company has been working on voice functionality for some time. In early April before GPT-4o was unveiled, OpenAI announced it built a new text-to-audio AI model, Voice Engine, though has yet to make it public.

In recent weeks rumors swirled that OpenAI will partner with Apple to integrate its AI technology into future iPhones. Speculation about an overhaul of Apple’s voice assistant Siri has been circulating ahead of this week’s Apple developers conference.