PrismML says it ran a 27B-parameter Qwen 3.6 model on an iPhone 17 Pro, bigger than any prior on-device model; sources: Apple held talks with PrismML about it (Aaron Tilley/The Information)