The US startup OpenAI has unveiled a new generative AI model that responds faster to questions and is capable of more natural interactions.

GPT-4o, released on Monday, features significantly upgraded text, visual and audio processing capabilities.

Notably, OpenAI says that unlike previous models, there is no lag in responding to audio, enabling it to converse more naturally with the user.

During a demonstration, one of the developers told the bot: "I'm doing a live demo, and frankly, I'm feeling a little bit nervous. Can you help me calm my nerves a little bit?"

The ChatGPT voice responded, "Just take a deep breath and remember, you're the expert."

The new AI model was also shown to function as a translator.

It translated a conversation between a person speaking in English and another speaking in Italian.

Competition to develop generative AI models has been intensifying.

Meta announced its latest technology last month. Google is expected to hold a developers' meeting on Tuesday, at which it is likely to unveil its newest product.