Apple Foundation Models are the language models Apple built to power Apple Intelligence: a small model that runs on the device and a larger one on Apple's servers.
In plain words
Apple says it protects privacy with on-device processing and Private Cloud Compute, and it trained two models to match: one of about 3 billion parameters that runs on the iPhone, iPad or Mac itself, and a bigger server model that runs on Apple's Private Cloud Compute. Both understand several languages and images and can call tools. In 2025 Apple opened the on-device model to outside apps through a Swift framework called Foundation Models.
A quick example
Imagine a note-taking app that suggests a title for each note. Using the Foundation Models framework, its developer could ask the on-device model for a title in a few lines of Swift, and the note never leaves the phone.
Where you'll meet it
- Apple Intelligence features across Apple devices.
- Third-party apps using the on-device model.
Related
Sources
- DocsUpdates to Apple's On-Device and Server Foundation Language Models, Apple Machine Learning Research · read 29 Sept 2026
- PaperApple Intelligence Foundation Language Models Tech Report 2025, Apple (arXiv) · read 29 Sept 2026
Checked by WhatIsAI against the sources above · 29 Sept 2026. Spotted an error? Report it.