Apple working to cram massive Gemini model into iPhone to power new Siri

Apple has reportedly delayed its AI-upgraded Siri and is now planning to use Google’s Gemini model later this year.

The new assistant is expected to blend on-device and cloud processing, with significant reliance on Google and Nvidia infrastructure.

That approach suggests Apple’s next major Siri leap may depend less on the iPhone’s Neural Engine and more on cloud AI.

The shift could also complicate Apple’s privacy-first message, as large AI models may be too demanding for fully local processing.