OpenAI consolidates teams to build audio model as first step toward audio-focused devices

OpenAI plans to announce a new audio language model in the first quarter of 2026 and views that model as an intentional step toward an audio-based physical hardware device, according to a report in The Information citing current and former employees.

The company has combined teams across engineering, product, and research into a single initiative to improve audio models, which researchers say lag behind text models in accuracy and speed.

Few ChatGPT users opt for the voice interface, with most preferring text. OpenAI hopes better audio models will shift user behavior toward voice and allow deployment in a wider range of devices, including cars.

OpenAI plans to release a family of physical devices in the coming years, starting with an audio-focused product. Employees have discussed various forms, such as smart speakers and glasses, with an emphasis on audio interfaces rather than screen-based ones.

Key Topics

AI, United States, Openai, Voice, Audio, Hardware, Devices