At WWDC 2025, Apple unveiled its Foundation Models framework, enabling developers to incorporate Apple’s on-device AI models to power new app functionalities.
Apple emphasized that this technology lets developers utilize AI models without incurring inference expenses. The local models also come with features like guided generation and built-in tool integration.
With iOS 26 now available to everyone, developers are enhancing their apps with features powered by Apple’s on-device AI. Compared to major models from OpenAI, Anthropic, Google, or Meta, Apple’s models are relatively compact. As a result, these on-device features primarily offer quality-of-life improvements rather than drastically transforming app workflows.
Here are several early examples of apps making use of Apple’s AI framework.
Lil Artist
Lil Artist provides a range of interactive activities to help children develop skills in creativity, math, and music. Developer Arima Jain added an AI-powered story creation feature with the iOS 26 release. Users can pick a character and theme, and the app will craft a story using AI. According to the developer, the story text is generated by the device’s local model.
Image Credits:Lil Artist
Daylish
Daylish’s developer is experimenting with a feature that automatically suggests emojis for timeline events in the daily planner, based on the event title.
MoneyCoach
MoneyCoach, a personal finance tracker, leverages local models for two useful functions: it provides spending insights—like identifying above-average grocery expenses for a week—and automatically recommends categories and subcategories for new expenses to speed up data entry.
Image Credits:MoneyCoach
LookUp
LookUp, a vocabulary learning app, utilizes Apple’s AI models to introduce two new modes. One mode generates examples for a word using the local model, and then prompts users to explain how that word is used in a sentence.
Image Credits:LookUp
Additionally, the app uses on-device models to create a visual map showing the origin of a word.
Image Credits:Lookup
Tasks
The Tasks app, like several others, uses local models to automatically recommend tags for entries. It also detects and schedules recurring tasks, and allows users to dictate items, which the app then converts into individual tasks offline.
Image Credits:Tasks
Day One
Day One, the journaling app owned by Automattic, employs Apple’s AI models to summarize entries and suggest titles. It also generates prompts that encourage users to elaborate further on their writing.
Image Credits:Day One
Crouton
Crouton, a recipe management app, utilizes Apple Intelligence to recommend tags for recipes, assign names to timers, and transform blocks of text into step-by-step cooking instructions using AI.
SignEasy
SignEasy, an app for digital signatures, leverages Apple’s local models to summarize contracts and highlight important points in the documents users are about to sign.
We’ll keep this list updated as we find more apps that utilize Apple’s local AI models.
