Foundation Models Framework Lab
A practical lab for building, testing, and evaluating apps with Apple's Foundation Models framework.
Foundation Lab is a native iOS and macOS workbench for learning, testing, and shipping with Apple's Foundation Models framework. It keeps the prompt, configuration, tools, transcript, and run evidence in one place while still providing focused labs for APIs that need a specialized interface. The project is written primarily in Swift, distributed under the MIT License license, first published in 2025. It has gained significant community traction with 1,124 stars and 68 forks on GitHub. Key topics include: ai, apple-foundation-models, apple-intelligence, foundation-models, foundation-models-framework.
Foundation Lab
Foundation Lab is a native iOS and macOS workbench for learning, testing, and
shipping with Apple's Foundation Models framework. It keeps the prompt,
configuration, tools, transcript, and run evidence in one place while still
providing focused labs for APIs that need a specialized interface.
The app is designed for two complementary workflows:
- Beginners can open a working recipe, change one thing, and run it immediately.
- Experienced developers can compose custom prompts and tools, inspect every run,
compare adapters, and use the repository's CLI and evaluation tooling.
App Structure
Foundation Lab has three primary destinations:
| Destination | Purpose |
|---|---|
| Library | Browse 18 editable recipes, 14 guided labs, three workshops, saved experiments, and two expert workspaces. |
| Playground | Edit prompts and instructions, configure the model and tools, stream responses, use voice input, save experiments, and export Swift. |
| Runs | Inspect persisted run status, configuration, transcript events, tool calls, timing, and token usage. |
Library entries identify how they open:
- Recipe opens in Playground and can be edited, run, and saved.
- Guided Lab uses a focused interface for a specific Foundation Models API.
- Workshop groups related schema, language, or Xcode 27 examples without adding another top-level destination.
- Workspace opens an expert tool such as Adapter Comparison or AppBench.
Requirements
- iOS 26.0+ or macOS 26.0+
- Apple Silicon for on-device model execution
- Apple Intelligence enabled for live model runs
- Xcode 26.6 or Xcode 27
The project builds with both Xcode 26.6 and Xcode 27. APIs introduced with the
OS 27 SDK are compiler- and availability-gated, so the core app remains usable
with Xcode 26 while Xcode 27 exposes the newest labs.
Getting Started
bashgit clone https://github.com/rudrankriyam/Foundation-Models-Framework-Lab.git cd Foundation-Models-Framework-Lab open FoundationLab.xcodeproj
Build from the command line:
bashxcodebuild \ -project FoundationLab.xcodeproj \ -scheme 'Foundation Lab' \ -destination 'generic/platform=macOS' \ CODE_SIGNING_ALLOWED=NO \ build xcodebuild \ -project FoundationLab.xcodeproj \ -scheme 'Foundation Lab' \ -destination 'generic/platform=iOS Simulator' \ CODE_SIGNING_ALLOWED=NO \ build
Live model execution requires a compatible physical device. Simulator builds
remain useful for compilation and interface validation.
Capabilities
Experiments and conversations
- Streaming multi-turn conversations with context-window management
- Editable instructions, sampling, response limits, runtime, and reasoning controls
- Saved experiment configurations and persistent run history
- Swift export for Playground configurations
- Speech recognition and synthesis integrated into Playground
Built-in tools
Nine ready-made tool recipes use the shared FoundationModelsTools package:
- Weather through Open-Meteo
- Keyless Search1 web search
- Contacts
- Calendar
- Reminders
- Location and place search
- Authorized HealthKit data
- Apple Music
- Web metadata
Tool recipes open in Playground, where tools can be combined or removed. Tools
that can change user data require confirmation through the app-owned workflow.
Structured output and applied projects
@Generablemodels and@Guideconstraints- Dynamic schemas, nested objects, unions, forms, and invoice extraction
- Multilingual sessions and supported-language inspection
- RAG document indexing and semantic retrieval with LumoKit and VecturaKit
- A HealthKit dashboard and chat grounded only in authorized Health data
Xcode 27 labs
When built with Xcode 27, Foundation Lab also demonstrates:
PrivateCloudComputeLanguageModel- Shared
LanguageModelexecution - Image attachments and references
- Explicit tool-calling modes
- Dynamic profiles and reasoning controls
- Transcript inspection and history transforms
- Context-budget visualization
- Custom model executors, including a video-capable provider bridge
The image-input probe under Tools/ImageInputProbe
can measure the current SDK's practical decoded-buffer boundary.
Expert Workspaces
Adapter Comparison
On macOS, import a .fmadapter package and run the same prompt through fresh
base-model and adapter sessions. The workspace shows both streams and diagnostic
time-to-first-token and total-duration measurements.
Training and export remain in the companion fmas CLI:
bashpython3.11 -m venv .venv-fmas source .venv-fmas/bin/activate python -m pip install -e Tools/AdapterStudio fmas init fmas setup fmas train-adapter --help fmas export --help
See Tools/AdapterStudio for the full workflow.
AppBench
AppBench is the repository's repeatable quality, safety, and performance suite.
The in-app workspace explains the protocol and artifacts; canonical results come
from the CLI on Mac or the signed device runner on physical iPhone and iPad hardware.
bashswift run appbench list swift run appbench --suite quick --model on-device swift run appbench --suite full --warmups 5 --repetitions 20 \ --json Tools/AppBench/Results/run.json \ --markdown Tools/AppBench/Results/run.md
See Tools/AppBench for workloads, methodology, graders, and
the device runner.
Command-Line Interface
The afm CLI uses the same FoundationLabCore and FoundationModelsKit
runtime as the app.
bashbrew tap rudrankriyam/tap brew install afm swift run afm --help swift run afm model status swift run afm session respond --prompt "Summarize Foundation Models."
See Tools/AFMCLI/README.md for the command reference.
AFM releases use afm-vx.y.z tags so CLI releases remain independent from app releases.
Repository Map
| Surface | Location | Purpose |
|---|---|---|
| Foundation Lab | Foundation Lab | Native Library, Playground, Runs, guided labs, and expert workspaces |
| FoundationLabCore | FoundationLabCore | UI-independent requests, results, use cases, providers, and experiment models |
| FoundationModelsKit | Packages/FoundationModelsKit | Transcript, context, history, and system-tool packages |
| AFM CLI | Tools/AFMCLI | Scriptable Foundation Models workflows |
| AppBench | Tools/AppBench | Quality, safety, and performance evaluation |
| Adapter tooling | Tools/AdapterStudio | Adapter training and export with fmas |
| Book playgrounds | BookPlaygrounds | Chapter-oriented #Playground examples |
The former standalone CLI, AppBench, and Adapter Studio repositories are
archived in favor of this shared implementation.
Swift Package Products
FoundationModelsKitprovides transcript history transforms, token estimation,
and context-budget utilities.FoundationModelsToolsprovides calendar, contacts, health, location, music,
reminders, weather, web search, and web metadata tools.FoundationLabCoreprovides the shared capability and experiment runtime.AppBenchCoreprovides the evaluation corpus, graders, runner, metrics, and reports.BenchmarkCoreremains as an AppBench compatibility alias.appbenchis the canonical benchmark executable.
Localization and Permissions
The app ships English, German, Spanish, French, Italian, Japanese, Korean,
Portuguese (Brazil), Simplified Chinese, and Traditional Chinese localizations.
Features request permissions only when needed. Depending on the selected recipe
or lab, the app may request microphone, speech recognition, contacts, calendar,
reminders, location, HealthKit, or Apple Music access.
Validation
bashswiftlint lint --strict --config .swiftlint.yml swift test
CI additionally builds Foundation Lab for macOS and iOS Simulator and validates
the AFM, AppBench, Adapter Studio, and TestFlight workflows.
TestFlight
Join the Foundation Lab beta on
TestFlight.
Pushes to main that affect the app can run the repository-local ASC workflow
in .asc/workflow.json through
foundation-lab-testflight.yml.
Agent Skills
The repository includes two reusable skills:
foundation-models-app-builderfor production Foundation Models patternsfoundation-models-os27-updaterfor Xcode 27 and OS 27 migrations
bashnpx skills add rudrankriyam/Foundation-Models-Framework-Lab \ --skill foundation-models-app-builder npx skills add rudrankriyam/Foundation-Models-Framework-Lab \ --skill foundation-models-os27-updater
Contributing
Contributions are welcome. Please open an issue or pull request with a focused
change and include the relevant lint, test, and build results.
License
Foundation Lab is available under the MIT License. See LICENSE.
Contributors
Showing top 9 contributors by commit count.
