Developers are actively integrating Apple's local artificial intelligence (AI) models into their applications following the rollout of iOS 26. This adoption leverages the Foundation Models framework, which Apple introduced earlier this year at WWDC 2025, providing developers access to on-device AI capabilities without incurring inference costs.
The framework offers features such as guided generation and tool calling, allowing for the creation of new functionalities within applications. While Apple's models are characterized as smaller compared to leading large language models from companies like OpenAI, Anthropic, Google, and Meta, their on-device processing capabilities are primarily enhancing 'quality of life' features rather than introducing fundamental changes to application workflows.
A range of applications has begun implementing these AI-powered features. The Lil Artist app now incorporates an AI story creator, enabling users to generate narratives based on selected characters and themes, with text generation powered by the local model, according to its developer. Daylish, a daily planner app, is prototyping automatic emoji suggestions for timeline events based on their titles. Finance tracking app MoneyCoach has introduced features for spending insights, such as comparisons to average spending, and automatic categorization suggestions for rapid entry.
Word learning application LookUp has added new modes utilizing Apple's AI, including a learning mode that generates word examples and prompts users to explain word usage. The app also uses on-device models to produce map views of word origins. Tasks, a to-do list application, implements local models for suggesting tags, detecting and scheduling recurring tasks, and transcribing spoken input into actionable tasks without requiring internet connectivity.
Automattic-owned journaling app Day One uses Apple's models to generate entry highlights, suggest titles, and create prompts to encourage deeper writing. Recipe manager Crouton employs Apple Intelligence for suggesting recipe tags, assigning names to timers, and converting blocks of text into sequential cooking steps. Digital signing app SignEasy leverages local models to extract key insights and provide summaries of contracts. Background sound application Dark Noise now allows users to describe a soundscape in words and generate one, with adjustable elements. The new F1 season tracking app, Lights Out, utilizes on-device AI to summarize race commentary.
Further applications demonstrating this integration include note-taking app Capture, which offers category suggestions as users type; sun and weather tracking app Lumy, providing weather-related suggestions; and Cardpointers, which enables users to ask questions about credit cards and offers. Guitar learning app Guitar Wiz explains chords, offers advanced player insights based on time intervals, and supports over 15 languages using the Apple Foundation Model framework.
