Apple in talks with PrismML to shrink AI models for iPhone, and crypto should pay attention

Running a 27-billion-parameter AI model on a phone sounds like the kind of claim that gets you laughed out of a pitch meeting. PrismML did it anyway.

Apple is in active discussions with the Caltech spinoff to explore whether its model-compression technology can bring significantly larger AI models directly onto iPhones, according to a report from The Information. The talks center on PrismML’s ability to shrink massive language models to a fraction of their original size while preserving their reasoning capabilities.

What PrismML actually built

PrismML compressed Alibaba’s Qwen 3.6 model, a 27-billion-parameter model that normally weighs in at roughly 54 GB, down to less than 4 GB. The shrunken model ran fully functional on an iPhone 17 Pro.

The company uses proprietary mathematical techniques including 1-bit and ternary weight architectures developed at Caltech. CEO Babak Hassibi, a professor of electrical engineering at the university, leads the operation. PrismML emerged from stealth on March 31, 2026, and has already secured $16.25 million in seed funding led by Khosla Ventures, with participation from Cerberus Ventures and Caltech itself.

PrismML has also shipped a consumer-facing product called Bonsai Studio, a mobile app that handles diffusion-based image generation entirely on-device. No internet required. The app launched by late May 2026.

The company claims its compression techniques enable on-device chat, reasoning, and coding tasks. Their pitch is that this could “fundamentally change the economics of AI.”

Why Apple cares, and why crypto should too

Apple’s interest makes strategic sense on multiple levels. The company has long positioned privacy as a core differentiator, and on-device AI processing is the logical endpoint of that philosophy. If you never send your data to a cloud server, there’s nothing to intercept, no third-party inference costs, and no latency from round-tripping to a data center.

Most “AI features” on phones currently shuttle requests to massive cloud infrastructure. PrismML’s technology, if it scales the way Apple hopes, would make the phone the actual computer again.

The competitive landscape and what to watch

Apple isn’t the only company chasing on-device AI. Google has been pushing its Gemini Nano models for Android, and Qualcomm has invested heavily in on-device inference through its Snapdragon processors. But the compression ratios PrismML is demonstrating, shrinking a model by roughly 93% while retaining functionality, would represent a step change in what’s possible.

The $16.25 million seed round is modest by AI startup standards. The Khosla Ventures backing adds credibility, but the real validation will come from whether Apple moves beyond exploratory talks to an actual integration or acquisition.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.