Verified80% confidenceFactTime unknown
The small language model running on-device for Apple Intelligence has approximately 3 billion parameters.
7
Sources
80%
Confidence
Long-term
Relevance
6/3/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedApple Intelligence was announced at WWDC 2024 and adopts a hybrid on-device/cloud inference strategy, running a small language model of approximately 3 billion parameters on-device.85% similarVerifiedApple Intelligence relies on Apple's self-developed Neural Engine to run a small language model of approximately 3 billion parameters on-device for local tasks, while accessing larger-scale models via a private cloud computing architecture in the cloud.83% similarUnverifiedApple devices can typically only run AI models with fewer than 3 billion parameters on-device.76% similarUnverifiedApple Intelligence's local model architecture is stored in a layered manner on macOS, including a base language model, adapter files, and various specialized models.75% similarUnverifiedApple launched an on-device model in the new version of iOS with a total parameter count of 20 billion, dynamically activating approximately 100 million to 400 million parameters per inference, and has demonstrated it running on the latest iPhone Pro.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/39220API
curl https://kongchang.com/api/v1/knowledge/claims/39220MCP
get_claim(id=39220)