This new feature enables more fluid, real-time conversations by allowing users to interrupt ChatGPT and respond to their emotions more intuitively.

On Tuesday, OpenAI announced via X that it is beginning the rollout of an advanced voice mode to a select group of ChatGPT Plus users.

The company postponed the launch of its realistic voice conversation feature from late June to July, citing the need for additional time to meet its quality standards.

The new audio features will enable users to converse with ChatGPT in real time, receiving instant responses and interrupting it while it speaks—key elements of natural conversation that have historically been difficult for AI assistants to achieve.

In June, OpenAI announced it was enhancing the model's ability to identify and reject certain content, while also focusing on improving user experience and preparing its infrastructure for scaling the model.

The company has been developing new generative AI products as it aims to stay ahead in the rapidly growing AI sector, where businesses are quickly adopting the technology.