GitHub - apple/python-apple-fm-sdk: Python bindings for access to the on-device model at the core of Apple Intelligence through the Foundation Models framework · GitHub

Source: original

Foundation Models SDK for Python

Python bindings for Apple's Foundation Models framework, providing access to the on-device foundation model at the core of Apple Intelligence on macOS.

Overview

The Foundation Models SDK for Python provides a Pythonic interface to Apple's Foundation Models framework.

You can:

• Evaluate Swift Foundation Models app features by running batch inference and analyzing results from Python
• Perform on-device inference with the system foundation model
• Stream real-time text generation responses
• Use guided generation with structured output schemas and constraints
• Get type-safe responses using Python decorators for guided generation
• Configure custom model settings for different model options
• Process transcripts exported from Swift apps for quality analysis

Keep in mind that it's your responsibility to design AI experiences with care. To learn about practical strategies you can implement in code, check out: Improving the safety of generative model output and Apple's Human Interface Guidelines on Generative AI.

Requirements

• macOS 26.0+
• Download Xcode 26.0+ and agree to the Xcode and Apple SDKs agreement in the Xcode app.
• Python 3.10+
• Apple Intelligence turned on for a compatible Mac

Contributing

This project is not yet taking contributions. Stay tuned!

Installation

pip install apple-fm-sdk

Alternatively, you can use the development install instructions below.

Documentation

• Documentation page
• Code examples

Basic usage

import apple_fm_sdk as fm import asyncio

async def main():

Get the default system foundation model

model = fm.SystemLanguageModel()

Check if the model is available

is_available, reason = model.is_available() if is_available:

Create a session

session = fm.LanguageModelSession()

Generate a response

response = await session.respond("Hello, how are you?") print(f"Model response: {response}") else: print(f"Foundation Models not available: {reason}")

Run async function

asyncio.run(main())

You can also use guided generation to ask the model to generate an object from a specific Python class marked with the generable decorator:

import apple_fm_sdk as fm

@fm.generable # This decorator signals this type be generated by a model class Cat: name: str age:int = fm.guide("Age in years", range=(0, 20))

async def generate_cat():

Get the default system foundation model

model = fm.SystemLanguageModel()

Check if the model is available

is_available, reason = model.is_available() if is_available:

Create a session

session = fm.LanguageModelSession()

Generate a response of the type Cat

cat = await session.respond("Generate an adorable rescue cat", generating=Cat) print(f"Model response: {cat}") else: print(f"Foundation Models not available: {reason}")

Development Installation

If you need to modify the SDK or install from source:

1. Get the code

git clone https://github.com/apple/python-apple-fm-sdk cd python-apple-fm-sdk

1. (Optional but recommended) Make a virtual environment. Install uv (or your package manager of choice) then:

uv venv source .venv/bin/activate

1. Install the package locally in editable mode:

uv sync

1. After making any change, be sure to build the project again and test:

uv pip install -e . pytest