On July 10, it was reported by Seeking Alpha thatPrismML has compressed Alibaba’s open-source Qwen 3.6 large language model to run on the iPhone 17 Pro. This has been confirmed.
CNBC reports that Apple is indeed evaluating PrismML’s technology, which the startup says can shrink powerful AI models enough to run directly on an iPhone while using up to 15x less memory.
The breakthrough could help Apple make Siri faster and more private by keeping more AI processing on-device instead of sending requests to the cloud. CNBC says if PrismML’s claims hold up in real-world testing, the technology could reshape demand for memory and datacenter compute — though analysts say AI will still require plenty of chips.
PrismML is an AI startup and research lab spun out of Caltech that has pioneered commercially viable 1-bit and ternary (+1, 0, -1) neural network architectures. By storing model weights at extreme low-bit precision, PrismML shrinks AI models by order-of-magnitude factors without severely sacrificing reasoning or generative capabilities.
The reported breakthrough has attracted Apple’s attention as the company looks to expand its on-device AI capabilities. The milestone “reflects Apple’s broader push to run AI directly on devices rather than relying on cloud infrastructure,” according to TrendForce.
MacRumors notes that running larger AI models directly on iPhones would allow more Apple Intelligence features to operate on-device instead of through Apple’s Private Cloud Compute servers. This could help reduce Apple’s operating costs while further enhancing user privacy.
I hope you’ll help support Apple World Today by becoming a patron. Almost all our income is from Patreon support and sponsored posts. Patreon pricing ranges from $2 to $10 a month. Thanks in advance for your support.
