foundation-models-on-device

Generate text and extract structured data on-device with Apple's FoundationModels framework.

1|Updated May 12, 2026
One-click install
npx skills add https://github.com/Manvendra08/TradingBot --skill foundation-models-on-device-manvendra08
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: foundation-models-on-device
Source: https://github.com/Manvendra08/TradingBot/tree/main/_agent/skills/foundation-models-on-device
Command: npx skills add https://github.com/Manvendra08/TradingBot --skill foundation-models-on-device-manvendra08

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill eliminates the need for cloud-dependent AI services in iOS apps, solving user privacy concerns and offline functionality gaps by enabling all AI processing to run directly on the user's device.

Core Features & Use Cases

  • On-Device Text Generation: Generate or summarize text, answer user queries, and power AI assistants without sending data to external servers.
  • Structured Data Extraction: Use @Generable macros to pull organized, typed data like user profiles, event details, or form inputs from natural language prompts.
  • Custom Tool Calling: Build domain-specific AI actions such as recipe searches, data lookups, or internal app workflows that the model can invoke automatically.
  • Use Case: A fitness app can use this Skill to generate personalized workout plans from user goals, stream the plan in real time as it generates, and keep all user health data on the device.

Quick Start

Use the foundation-models-on-device skill to build an iOS feature that extracts structured cat profile data from natural language user input and displays it in the app interface.

Frequently Asked Questions about foundation-models-on-device

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run AI text generation on-device in iOS without sending data to the cloud?

On-device text generation in iOS lets you generate text, summarize content, and answer user queries directly on the device. This Skill uses Apple's FoundationModels framework to process AI workloads locally, eliminating cloud dependency and preserving user privacy.

How do I extract structured data from natural language input in an iOS app?

To extract structured data from natural language in iOS, you use the @Generable macros provided by the FoundationModels framework. This allows you to pull organized, typed data such as user profiles or event details directly from user prompts on-device.

Can I build custom tool calling for domain-specific actions with Apple FoundationModels?

Yes, you can build custom tool calling with Apple FoundationModels to execute domain-specific AI actions. This allows the model to automatically invoke internal app workflows like recipe searches or data lookups without requiring an external server.

What are the limitations of on-device AI with Apple's FoundationModels framework?

On-device AI with FoundationModels operates within a 4,096 token context window limit. It also requires Apple Intelligence device eligibility and availability checks, meaning processing capabilities are constrained by the local hardware specifications.

Does on-device AI for iOS 26 support real-time streaming for UI updates?

Yes, on-device AI for iOS 26 supports snapshot streaming for real-time UI updates. This mechanism allows your app to display generated text or data progressively as the FoundationModels framework processes the request locally.