OpenAI’s Smartphone Takes Form With Fresh Specs and a Faster Timeline

Photo credit: Notebookcheck
Fresh details emerged today about the smartphone OpenAI has under development. The company wants this device to let AI agents handle daily tasks directly instead of forcing users to navigate grids of apps. Analyst Ming-Chi Kuo updated his findings on the project. Mass production now looks set for the first half of 2027. OpenAI hopes to move 30 million units between 2027 and 2028. That volume would position the phone as a serious option in the premium market alongside devices like the iPhone 18 Pro and the Galaxy S27 series.

So far, the processor has been the star of the show, since they are supposedly using a modified version of MediaTek’s Dimensity 9600 built on TSMC’s 2 nanometer technology. It has two independent neural processing units built-in, which divide the AI’s work between vision and language tasks, and the image signal processor has been updated to offer the AI a better view of what’s going on through the camera lens.

Memory options are all about meeting the demands of having AI working at all times. LPDDR6 RAM will keep the bandwidth flowing smoothly, while UFS 5.0 storage will handle read and write rates, ensuring that you don’t experience lag during long chats with the AI.

Earlier, in late April, stories surfaced indicating that collaborations were in place. MediaTek and Qualcomm will handle the chip work, with Luxshare serving as the manufacturing partner. Despite the fact that the schedule is being hastened, everything remains the same. And it appears that Jony Ive’s design business may be helping to shape the appearance of OpenAI’s hardware ambitions.

Sam Altman has been outspoken about his belief that we need to rethink how we create operating systems and interfaces, and the phone will most likely run Android. The idea is to keep the foundations for calls, networks, and basic hardware drivers intact, but with the flexibility to customize to the needs of the AI, because, at the end of the day, full control is essential when it comes to getting AI agents to work properly, as they need to be able to tap into the camera, data, and permissions.

OpenAI already has hundreds of millions of users using chatbots like ChatGPT, and the plan is to bundle this phone with subscription services, giving consumers a device that is seamlessly linked with the tools they need to act on the information they get. It’s all about establishing a natural flow, not a forced connection. Google and Samsung are experimenting with agent features on their current phones using software alone, but OpenAI believes that designing everything from the ground up around these agents is the way to get seamless performance.
[Source]