Apple Eyes Startup That Squeezes AI Models Onto iPhones
Apple is in talks with PrismML, whose compression tech runs AI models using up to 15x less memory — a potential game-changer for on-device AI.
Apple wants AI that actually lives on your phone — not in some distant data center — and it may have found the key to making that happen. The company is reportedly in talks with PrismML, a startup that has cracked a way to shrink AI models down to a fraction of their normal size without gutting their usefulness.
Here's the number that matters: PrismML claims its compressed version of Alibaba's Qwen model runs on up to 15 times less memory than the original. That's not a rounding error. That's the difference between an AI feature that works offline on your iPhone and one that requires a server farm every time you tap a button.
Read more Cascade Partners Adds Healthcare M&A Veteran Aaron Newman →
For Apple, the stakes are real. The company has been playing catch-up in the AI race — Siri still feels like it's running on dial-up compared to what rivals are shipping. On-device AI is Apple's best strategic play: it's faster, more private, and doesn't depend on a cloud connection. PrismML's compression approach could hand Apple a genuine edge, letting it run more capable models on existing hardware without forcing users to upgrade.
This is the kind of quiet, infrastructure-level deal that doesn't grab headlines — but it's exactly where AI battles are actually won. If Apple closes this and bakes PrismML's tech deep into iOS, the ripple effects hit every developer building on Apple's platform and every competitor trying to match on-device performance. Watch this space.
Continue reading at US Top News and Analysis.
