Apple Eyes AI Startup That Squeezes Models Into iPhone Memory
Apple is in talks with PrismML, whose tech compresses AI models to use up to 15x less memory — a potential game-changer for on-device AI.
Apple is reportedly in conversations with PrismML, a startup that has figured out how to shrink large AI models small enough to run natively on an iPhone. If this deal moves forward, it could be the shot in the arm Apple Intelligence has been waiting for.
Here's the stat that matters: PrismML claims its compressed version of Alibaba's Qwen model uses up to 15 times less memory than the original. That's not a rounding error — that's the difference between running sophisticated AI on your phone versus needing a data center to do the heavy lifting.
Read more World Cup Semifinals Spark Late Travel Surge to U.S. Host Cities →
On-device AI is the competitive battleground right now. Running models locally means faster responses, no cloud dependency, and stronger privacy guarantees — things Apple has made central to its brand pitch. But cramming capable AI into a device with limited RAM and compute has been the hard part. PrismML's compression approach could be exactly the unlock Apple needs to close the gap with rivals.
Apple has been playing catch-up in the generative AI race, leaning on partnerships — most notably with OpenAI — to bolster Siri and Apple Intelligence features. A deal with PrismML would signal a shift toward owning more of that stack in-house, which is exactly how Apple likes to operate long-term. Control the silicon, control the software, control the AI.
Nothing is finalized, and early-stage talks can fall apart fast. But if Apple lands this, expect on-device AI capabilities to accelerate well beyond what current iPhones can do. Watch this space. Continue reading at US Top News and Analysis.
