Apple looks to shrink AI models for iPhones

https://www.businessreport.com/article/apple-looks-to-shrink-ai-models-for-iphones

Publish Date: 2026-07-14 15:29:00

Apple is reportedly in very early discussions with PrismML, a Silicon Valley startup spun out of the California Institute of Technology and backed by Khosla Ventures, about technology that could dramatically reduce the size of powerful AI models so they can run directly on iPhones, CNBC reports.

PrismML recently announced that it compressed Alibaba’s open-source 27 billion parameter Qwen model from roughly 54 GB to less than 4 GB, allowing all 27 billion parameters to run on an iPhone 15 or newer. According to the company, its compression method reduces memory usage by 10 to 15 times, increases processing speed by 6 to 8 times and lowers energy consumption by 3 to 6 times while sacrificing only a few percentage points of overall performance, with factual recall declining more than reasoning, math and coding abilities.

Apple is reportedly evaluating the technology’s speed, energy efficiency and on-device performance as part of its broader effort to improve Siri and expand on-device AI capabilities. Running more AI locally could deliver faster responses, strengthen user privacy by keeping more sensitive data on the device, reduce reliance on cloud computing and enable certain AI features to work without an internet connection.

Analysts say this approach could support more advanced applications, including computational photography, video generation and health and fitness tools that process personal information. However, they caution that PrismML’s claims must still be validated through large-scale, real-world testing, particularly with respect to battery life, reliability and performance across millions of devices.

If the technology performs as claimed, it could help Apple advance its AI strategy while reinforcing its emphasis on privacy and hardware-software integration. The breakthrough could shift a greater share of AI processing from data centers to personal devices, although experts note it is unlikely to reduce overall…

Source