Search
On July 10, 2026, Apple was reported to be collaborating with the startup PrismML to delve into the deployment of a 27-billion-parameter AI large-scale model on iPhones. PrismML has successfully achieved the compression of Alibaba's open-source Qwen 3.6 large-scale model, enabling it to operate seamlessly on the iPhone 17 Pro. The parameter scale of this model significantly surpasses that of Apple's current on-device models, with the capability to activate all parameters concurrently during operation. Apple's initiative is geared towards minimizing dependence on cloud computing resources, fortifying user privacy protection, and amplifying the device's local AI processing prowess. Presently, the collaboration between the two entities is still in its exploratory phase, and the timeline for the technical realization of this endeavor remains ambiguous.
-
C114 Communication Network -
Communication Home
