OpenAI is starting to roll out an advanced voice mode to a small group of ChatGPT Plus users, the Microsoft-backed artificial intelligence startup said on Tuesday in a post on X.

The company had delayed the roll-out of the realistic voice conversation experience to July from late-June, saying it needed time to reach its launch standard.

The new audio capabilities will allow users to speak to ChatGPT and receive real-time responses without delay, as well as interrupt ChatGPT while it is speaking - both tenets of realistic conversations that have proven to be a challenge for AI assistants.

OpenAI said in June it was improving the model's ability to detect and refuse certain content, while working on bettering the user experience and preparing its infrastructure to scale the model.

(For top technology news of the day, subscribe to our tech newsletter Today’s Cache)

The company has been working to introduce new generative AI products, as it seeks to maintain its edge in the booming AI race with businesses rushing to adopt the technology.