OpenAI may build a phone where AI agents replace apps
▼ Summary
– OpenAI is reportedly developing a smartphone with MediaTek, Qualcomm, and Luxshare, according to analyst Ming-Chi Kuo.
– The smartphone would rely on AI agents instead of apps, allowing OpenAI to bypass restrictions from Apple and Google’s app pipelines.
– The device is designed to continuously understand user context, giving OpenAI access to more data than a smartphone app could provide.
– The smartphone will use a mix of small on-device models and cloud models to handle different tasks.
– Specifications and component suppliers are expected to be finalized by late 2026 or early 2027, with mass production starting in 2028.
A new analyst note from industry insider Ming-Chi Kuo suggests that OpenAI may be developing a custom smartphone in partnership with key chipmakers MediaTek and Qualcomm, with Luxshare serving as the co-design and manufacturing partner. This would mark a significant hardware expansion for the company behind ChatGPT.
Kuo, known for accurately forecasting Apple’s product plans, indicates that the proposed device would likely eschew traditional app-based interfaces in favor of AI agents that handle tasks directly. This approach would bypass the current restrictions imposed by Apple and Google, who tightly control app distribution and system-level access. By owning the entire hardware stack, OpenAI could integrate its AI deeply into every phone function without external limitations.
The timing is strategic. With ChatGPT nearing a billion weekly active users, a dedicated hardware product could help OpenAI extend its reach into daily consumer life beyond just a software service. This vision aligns with broader industry sentiment. Vibe coding app developers are already predicting a future without apps, and Nothing CEO Carl Pei recently stated at SXSW that apps will eventually disappear.
Kuo’s analysis suggests the OpenAI smartphone would be engineered for continuous contextual awareness, learning user habits more intimately than any app could. To balance performance and privacy, the device would likely use a mix of on-device small models for quick tasks and cloud-based models for more complex requests. This hybrid approach could deliver a seamless, intelligent user experience.
According to the note, final specifications and component suppliers are expected to be locked in by the end of this year or early 2027, with mass production slated for 2028. This timeline follows earlier reports that OpenAI plans to announce its first hardware product in the second half of 2026, with several sources pointing to uniquely designed earbuds as that initial device.
OpenAI declined to comment on the story at the time of writing.
(Source: TechCrunch)
