For a while now, Open AI has been working on an AI gadget with built-in cameras, microphones and speakers that will be able to record what is happening in the environment and answer users’ questions.

According to The Information, the device will be mostly controlled by voice, which makes sense since the device has no screen.

Thus, the focus is currently on developing a new voice model that sounds more natural and is capable of speaking and understanding speech at the same time, which is not as easy as one might think.

The new voice model is expected to be ready within a few months, but the device itself will be launched in the fall at the earliest.

This article originally appeared on ComputerSweden.

More OpenAI news:

• OpenAI rolls out GPT-5.1 to refine ChatGPT with adaptive reasoning and personalization
• OpenAI launches GPT-5.2 as it battles Google’s Gemini 3 for AI model supremacy
• What does OpenAI’s ‘Code Red’ warning mean for Microsoft?