Tools · SPBoucher
Menu

Foundation Models (Apple)

by Apple Inc. · Framework · also known as Apple Foundation Models framework, FoundationModels

Apple's Swift framework giving apps access to the on-device Apple Intelligence language model with guided generation, tool calling and sessions, on OS 26 and later

ProprietaryFreeLocalCloud
Visit official site Compare

Link verified Nov 2, 2026

Overview

Foundation Models is the Swift framework Apple introduced with iOS 26, iPadOS 26, macOS 26 and visionOS 26 (watchOS 27) that lets apps call the large language models behind Apple Intelligence. The default is the on-device model, which runs with no network connection and no per-request cost; the framework also offers a dynamic profile API for selecting other model configurations, including Apple's Private Cloud Compute models or a server-side provider, when a task needs more reasoning or context.

Its headline feature is guided generation: by marking a Swift type with the @Generable macro (and @Guide annotations on fields) the developer gets strongly typed output rather than free text to parse. Developers define tools conforming to a Tool protocol that the model can call to look up data or trigger app actions, hold multi-turn conversations in a LanguageModelSession, stream partial results, and check availability before use because the model is present only on devices that support Apple Intelligence with the feature enabled.

The on-device model is small compared with frontier cloud models and is intended for summarization, extraction, classification, short creative text and in-app dialogue rather than open-ended knowledge or long reasoning. It is proprietary, Swift-only and limited to current Apple OS releases on capable hardware.

Capabilities

  • On-device LLM inference with no server dependency
  • Guided generation into Swift types via @Generable
  • Tool calling through a Tool protocol
  • Multi-turn sessions with instructions and streaming
  • Dynamic profile selection across on-device and cloud models
  • Availability checks for Apple Intelligence support

Typical workflows

  1. Summarize or tag user content inside an app offline
  2. Extract structured fields from text into a Swift struct
  3. Let the model call an app action through a tool

Strengths

  • No API keys, quotas or network for the default model
  • Typed output removes JSON parsing
  • Privacy by design: data stays on the device

Limitations

  • Requires OS 26 or later on Apple Intelligence-capable hardware
  • On-device model is small with limited context and knowledge
  • Swift and Apple platforms only

Prerequisites

  • Xcode 26 or later
  • Device or simulator with Apple Intelligence enabled

Who it is for

  • iOS and macOS developers adding language features
  • Teams needing offline, private text processing

Alternatives & comparison

  • Claude Platform

    Anthropic's developer platform for the Claude models: Messages API, tool use, structured outputs, prompt caching, batch processing and Managed Agents, plus cloud availability

    ProprietaryUsage-based
  • Core ML

    Apple's on-device machine-learning runtime for iOS and macOS apps, with the open-source coremltools package for converting PyTorch and other models to its format

    MixedFree
  • MLX LM

    Python package and CLI for running, serving, quantizing and fine-tuning large language models on Apple Silicon with MLX

    Open sourceOpen source
  • Ollama

    Open-source runtime and model manager for running open-weight LLMs locally on macOS, Windows and Linux, with a REST API, OpenAI- and Anthropic-compatible endpoints and optional cloud models

    Open sourceOpen source
  • OpenAI API

    OpenAI's hosted developer platform: Responses and Chat Completions APIs, Realtime voice, embeddings, images, audio, fine-tuning, evals and managed agents

    ProprietaryUsage-based
Compare Foundation Models (Apple) with its alternatives →

Sources & verification

Primary sources consulted for this record. Claims are tied to the fields each source supports; the automated link checker re-verifies every URL on a schedule.

Confidence
High
Reviewed
Nov 1, 2026
Links
Verified · Nov 2, 2026

Spotted an error?

Broken link, outdated license, wrong category, or a tool that no longer exists? Tell us; reports are reviewed by the editor.

llama.cpp

ggml-org (Georgi Gerganov and contributors) · Local LLM inference & serving

In-depth entry

C/C++ inference engine for LLMs and vision-language models with GGUF quantized weights, Metal, CUDA, Vulkan and other backends, a CLI and an OpenAI-compatible server

MITOpen sourceFreeLocalSelf-hostedNov 2, 2026

LM Studio

Element Labs, Inc. · Local LLM inference & serving

In-depth entry

Proprietary desktop app for discovering, downloading and running local LLMs with llama.cpp and MLX engines, a chat UI, an OpenAI-compatible server, CLI, SDKs and a headless daemon

ProprietaryFreemiumPaidLocalBroken

MLX

Apple Inc. · Apple Silicon ML & MLX

In-depth entry

Apple's open-source array framework for machine learning on Apple Silicon, with NumPy-like Python, C++ and Swift APIs and unified CPU/GPU memory

MITOpen sourceFreeLocalNov 2, 2026

Metal

Apple Inc. · Apple Silicon ML & MLX

Apple's low-level GPU programming framework for graphics and parallel compute on iOS, macOS and visionOS, the layer beneath MLX, Core ML and most Apple Silicon ML runtimes

ProprietaryFreeLocalNov 2, 2026

Create ML

Apple Inc. · Apple Silicon ML & MLX

Apple's app and Swift framework for training image, text, sound, tabular and activity models on a Mac without writing training code, exporting directly to Core ML

ProprietaryFreeLocalNov 2, 2026

exo

Exo Labs · Distributed & parallel computing

Open-source software that pools several Apple Silicon Macs (and Linux hosts) into one local inference cluster, sharding large models across devices with MLX

Apache-2.0Open sourceFreeLocalSelf-hostedNov 2, 2026