I've been successfully integrating the Foundation Models framework into my healthcare app using structured generation with @Generable schemas. While my initial testing (20-30 iterations) shows promising results, I need to validate consistency and reliability at scale before production deployment. Question Is there a recommended approach for automated, large-scale testing of Foundation Models responses? Specifically, I'm looking to: Automate 1000+ test iterations with consistent prompts and structured schemas Measure response consistency across identical inputs Validate structured output reliability (proper schema adherence, no generation failures) Collect performance metrics (TTFT, TPS) for optimization Specific Questions Framework Limitations: Are there any undocumented rate limits or thermal throttling considerations for rapid session creation/destruction? Performance Tools: Can Xcode's Foundation Models Instrument be used programmatically, or only through Instruments UI? Automation Integration: Any recommendations for integrating with testing frameworks? Session Reuse: Is it better to reuse a single LanguageModelSession or create fresh sessions for each test iteration? Use Case Context My wellness app provides medically safe activity recommendations based on user health profiles. The Foundation Models framework processes health context and generates structured recommendations for exercises, nutrition, and lifestyle activities. Given the safety implications of providing health-related guidance, I need rigorous validation to ensure the model consistently produces appropriate, well-formed recommendations across diverse user scenarios and health conditions. Has anyone in the community built similar large-scale testing infrastructure for Foundation Models? Any insights on best practices or potential pitfalls would be greatly appreciated.

I installed Xcode 26.0 beta and downloaded the generative models sample from here: https://developer.apple.com/documentation/foundationmodels/adding-intelligent-app-features-with-generative-models But when I run it in the iOS 26.0 simulator, I get the error shown here. What's going wrong?

Hi, I have an app that uses Core Data to store user information and display it in various views. I want to know if it's possible to easily integrate this setup with FoundationModels to make it easier for the user to query and manipulate the information, and if so, how would I go about it? Can the model be pointed to the database schema file and the SQLite file sitting in the user's app group container to parse out the information needed? And/or should the NSManagedObjects be made @Generable for better output? Any guidance about this would be useful.

Was just wondering why the foundation model documentation is no longer available, thanks! https://developer.apple.com/documentation/FoundationModels

On macOS Tahoe26.0, iOS 26.0 (23A5287g) not emulator, Xcode 26.0 beta 3 (17A5276g) Follow this tutorial Testing your asset packs locally The start the test server command I use this command line to start the test server：xcrun ba-serve --host 192.168.0.109 test.aar The terminal showThe content displayed on the terminal is： Loading asset packs… Loading the asset pack at “test.aar”… Listening on port 63125…… Choose an identity in the panel to continue. Listening on port 63125… running the project, Xcode reports an error:Download failed: Could not connect to the server. I use iPhone safari visit this website: https://192.168.0.109:63125, on the page display "Hello, world!" There are too few error messages in both of the above questions. I have no idea what the specific reasons are.I hope someone can offer some guidance. Best Regards. { "assetPackID": "testVideoAssetPack", "downloadPolicy": { "prefetch": { "installationEventTypes": ["firstInstallation", "subsequentUpdate"] } }, "fileSelectors": [ { "file": "video/test.mp4" } ], "platforms": [ "iOS" ] } this is my Manifest.json

In WWDC25 Metal 4 released quite excited new features for machine learning optimization, but as we all know the pytorch based on metal shader performance (mps) is the one of most important tools for Mac machine learning area.but on mps introduced website we cannot see any support information for metal4.

Hi Apple team, When using AppShortcutsProvider, I hit the hard limit: Each app may have at most 10 App Shortcuts. This feels limiting for apps that offer multiple workflows and would benefit from deeper Siri integration. Could this cap be raised — ideally to 30 — to support broader use of AppIntents, enhance Siri automation, and unlock more system-level capabilities? AppShortcuts are a fantastic tool. Increasing the limit would make them even more powerful. Thanks!

Introduced in the Keynote was the 3D Lock Screen images with the kangaroo: https://9to5mac.com/wp-content/uploads/sites/6/2025/06/3d-lock-screen-2.gif I can't see any mention on if this effect is available for developers with an API to convert flat 2D photos in to the same 3D feeling image. Does anyone know if there is an API?

I'm trying to use the Spatial model to perform Object Tracking on a .usdz file that I create. After loading the file, which I can view correctly in the console, I start the training. Initially, I notice that the disk usage on my PC increases. After several GB, the usage stops, but the training progress remains for hours at 0.00% with the message "About 8hr." How can I understand what the issue is? Has anyone else experienced the same problem? Thanks Diego

I downloaded Xcode Beta 1 on my mac (did not upgrade the OS). The target OS level of iOS26 and the device simulator for iOS26 is downloaded and selected as the target. When I try a simple Playground in Xcode ( #Playground ) I get a session error. #Playground { let avail = SystemLanguageModel.default.availability if avail != .available { print("SystemLanguageModel not available") return } let session = LanguageModelSession() do { let response = try await session.respond(to: "Create a recipe for apple pie") } catch { print(error) } } The error I get is: Asset com.apple.gm.safety_deny_input.foundation_models.framework.api not found in Model Catalog Is there a way to test drive the FoundationModel code without upgrading to macos26?

Problem: We trained a LoRA adapter for Apple's FoundationModels framework using their TAMM (Training Adapter for Model Modification) toolkit v0.2.0 on macOS 26 beta 4. The adapter trains successfully but fails to load with: "Adapter is not compatible with the current system base model." TAMM 2.0 contains export/constants.py with: BASE_SIGNATURE = "9799725ff8e851184037110b422d891ad3b92ec1" Findings: Adapter Export Process: In export_fmadapter.py def write_metadata(...): self_dict[MetadataKeys.BASE_SIGNATURE] = BASE_SIGNATURE # Hardcoded value The Compatibility Check: - When loading an adapter, Apple's system compares the adapter's baseModelSignature with the current system model - If they don't match: compatibleAdapterNotFound error - The error doesn't reveal the expected signature Questions: - How is BASE_SIGNATURE derived from the base model? - Is it SHA-1 of base-model.pt or some other computation? - Can we compute the correct signature for beta 4? - Or do we need Apple to release TAMM v0.3.0 with updated signature?

使用MPS来加速机器学习功能，有时是否与torch会有适配性问题？

Hey guys 👋 I’ve been thinking about a feature idea for iOS that could totally change the way we interact with apps like Twitter/X. Imagine if we could define our own recommendation algorithm, and have an AI on the iPhone that replaces the suggested tweets in the feed with ones that match our personal interests — based on public tweets, and without hacking anything. Kinda like a personalized "AI skin" over the app that curates content you actually care about. Feels like this would make content way more relevant and less algorithmically manipulative. Would love to know what you all think — and if Apple could pull this off 🔥

Hello, We find that models sometimes load very fast (<< 1 second) and sometimes encounter very long load times (>> 120 seconds). During such slow load times, the model is being compiled. We would greatly appreciate the ability to check cache validity via CoreML and determine that we are about to encounter long load times so that we can mitigate and provide a good user experience. A secondary issue: sometimes the cache is corrupted (typically .mpsgraphpackage yielding Metal cold asserts). This yields load failures and OS errors that persist between launches, and we have to manually nuke the cache (~/Library/..../my-app/...) for the CoreML assets. A CoreML API for clearing caches and hardening from asserts across the load paths would be appreciated

iOS26 is supported by a wider range of devices than are able to run AI, e.g iPhone 12 runs iOS26, but does not support AI. How do we determine in code if AI is supported on a device ? How do we determine what features use AI under the hood ? Thanks, Steve.

I am experimenting with Foundation Models in my time tracking app to analyze users tracked events, but I am finding that the model struggles with even basic computation of time. Specifically converting from seconds to hours and minutes. To give just one example, when I prompt: "Convert 3672 seconds to hours, minutes, and seconds. Don't include the calculations in the resulting output" I get this: "3672 seconds is equal to 1 hour, 0 minutes, and 36 seconds". Which is clearly wrong - it should be 1 hour, 1 minute, and 12 seconds. Another issue that I saw a lot is that seconds were considered to be minutes, or that the hours were just completely off. What can I do to make the support for math better? Or is that just something that the model is not meant to be used for?

Is foundation models matured enough to take input from the Apple Vision framework to generate responses? Something similar to what google's gemini does although in a much smaller scale and for a very specific niche.

I've checked on pypi.org and it appears to only have arm64 packages, has x86 with AMD been deprecated?

Hello, I posted an issue on the coremltools GitHub about my Core ML models not performing as well on iOS 17 vs iOS 16 but I'm posting it here just in case. TL;DR The same model on the same device/chip performs far slower (doesn't use the Neural Engine) on iOS 17 compared to iOS 16. Longer description The following screenshots show the performance of the same model (a PyTorch computer vision model) on an iPhone SE 3rd gen and iPhone 13 Pro (both use the A15 Bionic). iOS 16 - iPhone SE 3rd Gen (A15 Bioinc) iOS 16 uses the ANE and results in fast prediction, load and compilation times. iOS 17 - iPhone 13 Pro (A15 Bionic) iOS 17 doesn't seem to use the ANE, thus the prediction, load and compilation times are all slower. Code To Reproduce The following is my code I'm using to export my PyTorch vision model (using coremltools). I've used the same code for the past few months with sensational results on iOS 16. # Convert to Core ML using the Unified Conversion API coreml_model = ct.convert( model=traced_model, inputs=[image_input], outputs=[ct.TensorType(name="output")], classifier_config=ct.ClassifierConfig(class_names), convert_to="neuralnetwork", # compute_precision=ct.precision.FLOAT16, compute_units=ct.ComputeUnit.ALL ) System environment: Xcode version: 15.0 coremltools version: 7.0.0 OS (e.g. MacOS version or Linux type): Linux Ubuntu 20.04 (for exporting), macOS 13.6 (for testing on Xcode) Any other relevant version information (e.g. PyTorch or TensorFlow version): PyTorch 2.0 Additional context This happens across "neuralnetwork" and "mlprogram" type models, neither use the ANE on iOS 17 but both use the ANE on iOS 16 If anyone has a similar experience, I'd love to hear more. Otherwise, if I'm doing something wrong for the exporting of models for iOS 17+, please let me know. Thank you!

While building an app with large language model inferencing on device, I got gibberish output. After carefully examining every detail, I found it's caused by the fused scaledDotProductAttention operation. I switched back to the discrete operations and problem solved. To reproduce the bug, please check https://github.com/zhoudan111/MPSGraph_SDPA_bug