Apple has found a “temporary” $1 billion-a-year solution to its AI problem: Google’s Gemini. The “ultrapowerful” 1.2 trillion parameter model will be the “behind-the-scenes” engine for the new “Linwood” Siri, set for a spring release.
This “interim solution” is part of Apple’s “Glenwood” project. Google’s AI was chosen after a “bake-off” where it beat OpenAI and Anthropic, and it will handle all complex “summariser” and “planner” functions.
The new Siri will be a hybrid, using Apple’s 150-billion parameter models for simple tasks and Google’s AI for heavy lifting. This is a reluctant admission of Apple’s AI lag, overseen by Craig Federighi and Mike Rockwell.
For Google, this is a major strategic win, positioning it as an “AI supplier.” For Apple, it’s a “temporary fix” as its teams race to build a 1T+ model replacement.
Privacy is non-negotiable. The Gemini model will be hosted on Apple’s “walled-off” Private Cloud Compute servers, ensuring Google gets its $1B fee but no access to user data.
