New Delhi: OpenAI has been reported to be gearing up to roll out the latest audio-focused GPT models as part of its broader push toward a first-generation hardware device that relies primarily on voice-based interaction. OpenAI’s maiden AI device is expected to be largely audio-driven with no display and a strong emphasis on conversational AI. OpenAI is upgrading its audio model architecture ahead of the hardware launch. These latest models are stated to generate more natural and expressive speech, deliver more detailed responses, and handle interruptions better than current versions.
The updated models are also expected to support overlapping speech, enabling the AI to speak while users are talking, a capability that existing ChatGPT voice features do not yet offer. OpenAI is reported to be aiming to release these latest audio models in the first quarter of 2026. OpenAI’s first consumer device is being developed in collaboration with former Apple design chief Jony Ive, with the project having moved into the prototyping phase in 2025.
The device is likely to be compact, screen-free, and designed to pick up audio and visual cues from its surroundings. OpenAI could rely on a small projector to display information on nearby surfaces instead of including a built-in display. OpenAI has discussed form factors such as smart glasses and speaker-style devices without a display. These ideas point toward a product designed to sit along with the users throughout the day, rather than replacing phones or laptops outright.
