Apple today announced a major overhaul of its Apple Intelligence platform, revealing a new architecture built on core models developed in collaboration with Google using the technologies underlying the Gemini family.
The new architecture centers around Apple Foundation models co-developed with Google, which Apple says are tailored to run both on-device and on servers through its existing Private Cloud Compute infrastructure. Apple described the collaboration as a “deep” one that it says unlocks what it calls a “huge upgrade” for Apple Intelligence, bringing cutting-edge understanding and reasoning capabilities as well as multimodal support including understanding and image generation.
The upgraded templates support new feature use cases, including realistic image creation, advanced photo editing, and visual question answering. Some devices will receive a more powerful version of the model with additional features including voice generation, improved dictation accuracy and better natural language understanding, although Apple did not specify which devices were eligible.
A new System Orchestrator sits at the center of the revised architecture, securely coordinating “Apple Intelligence” features across Apple platforms. Apple says the orchestrator allows the system to adapt its responses based on the user’s active application and current task, enabling what the company described as true system-wide intelligence.
Apple used this announcement to define its approach in contrast to its competitors, which it describes as “racing forward” without worrying about users. On privacy, the company reiterated that “Apple Intelligence” relies on on-device processing and Private Cloud Compute, with the promise that user data is only used to fulfill the immediate request and is not accessible to Apple or third parties. Apple added that external experts can verify these privacy guarantees “at any time.”
