WWDC25: Combine Metal 4 machine learning and graphics Demonstrated a way to combine neural network in the graphics pipeline directly through the shaders, using an example of Texture Compression. However there is no mention of using which ML technique texture is compressed. Can anyone point me to some well known model/s for this particular use case shown in WWDC25.

Hello everybody! I’m encountering an unexpected guardrailViolation error when using Foundation Models on macOS Beta 3 (Tahoe) with an Apple M2 Pro chip. This issue didn’t occur on Beta 1 or Beta 2 using the same codebase. Reproduction Context I’m developing an app that leverages Foundation Models for structured generation, paired with a local database tool. After upgrading to macOS Beta 3, I started receiving this error consistently, despite no changes in the generation logic. To isolate the issue, I opened the official WWDC sample project from the Adding intelligent app features with generative models and the same guardrailViolation error appeared without any modifications. Simplified Working Example I attempted to narrow down the issue by starting with a minimal prompt structure. This basic case works fine: import Foundation import Playgrounds import FoundationModels @Generable struct GeneableLandmark { @Guide(description: "Name of the landmark to visit") var name: String } final class LandmarkSuggestionGenerator { var landmarkSuggestion: GeneableLandmark.PartiallyGenerated? private var session: LanguageModelSession init(){ self.session = LanguageModelSession( instructions: Instructions { """ generate a list of landmarks to visit """ } ) } func createLandmarkSuggestion(location: String) async throws { let stream = session.streamResponse( generating: GeneableLandmark.self, options: GenerationOptions(sampling: .greedy), includeSchemaInPrompt: false ) { """ Generate a list of landmarks to viist in \(location) """ } for try await partialResponse in stream { landmarkSuggestion = partialResponse } } } #Playground { let generator = LandmarkSuggestionGenerator() Task { do { try await generator.createLandmarkSuggestion(location: "New york") if let suggestion = generator.landmarkSuggestion { print("Suggested landmark: \(suggestion)") } else { print("No suggestion generated.") } } catch { print("Error generating landmark suggestion: \(error)") } } } But as soon as I use the Sample ItineraryPlanner: #Playground { // Example landmark for demonstration let exampleLandmark = Landmark( id: 1, name: "San Francisco", continent: "North America", description: "A vibrant city by the bay known for the Golden Gate Bridge.", shortDescription: "Iconic Californian city.", latitude: 37.7749, longitude: -122.4194, span: 0.2, placeID: nil ) let planner = ItineraryPlanner(landmark: exampleLandmark) Task { do { try await planner.suggestItinerary(dayCount: 3) if let itinerary = planner.itinerary { print("Suggested itinerary: \(itinerary)") } else { print("No itinerary generated.") } } catch { print("Error generating itinerary: \(error)") } } } The error pops up: Multiline Error generating itinerary: guardrailViolation(FoundationModels.LanguageModelSession. >GenerationError.Context(debug Description: "May contain sensitive or unsafe content", >underlyingErrors: [FoundationModels. LanguageModelSession. Gene >rationError.guardrailViolation(FoundationMo dels. >LanguageModelSession.GenerationError.C ontext (debugDescription: >"May contain unsafe content", underlyingErrors: []))])) Based on my tests: The error may not be tied to structure complexity (since more nested structures work) The issue may stem from the tools or prompt content used inside the ItineraryPlanner The guardrail sensitivity may have increased or changed in Beta 3, affecting models that worked in earlier betas Thank you in advance for your help. Let me know if more details or reproducible code samples are needed - I’m happy to provide them. Best, Sasha Morozov

Just tried to write a very simple test of using foundation models, but it gave me the error like this "ModelManager received unentitled request. Expected entitlement com.apple.modelmanager.inference establishment of session failed with Missing entitlement: com.apple.modelmanager.inference" The simple code is listed below: let session: LanguageModelSession = LanguageModelSession() let response = try? await session.respond(to: "What is the capital of France?") print("Response: (response)") So what's the problem of this one?

Is there any way to ensure iOS apps we develop using Foundation Models can only be purchasable/downloadable on App Store by folks with capable devices? I would've thought there would be a Required Capabilities that App Store would hook into, but I don't seem to see it in the documentation here: https://developer.apple.com/documentation/bundleresources/information-property-list/uirequireddevicecapabilities The closest seems to be iphone-performance-gaming-tier as that seems to target all M1 and above chips on iPhone & iPad. There is an ipad-minimum-performance-m1 that would more reasonably seem to ensure Foundation Models is likely available, but that doesn't help with iPhone. So far, it seems the only path would be to set Minimum Deployment to iOS 26 and add iphone-performance-gaming-tier as a required capability, but I'm a bit worried that capability might diverge in the future from what's Foundation Model / Apple Intelligence capable. While I understand for the majority of apps they'll want to just selectively add in Apple Intelligence features and so can be usable by folks whose devices don't support it, the app experience I'm building doesn't make sense without the Foundation Models being available and I'd rather not have a large number of users downloading the app to be told "Sorry, you're not Apple Intelligence capable"

I want to use Foundation Models in a project, but I know my users will want to avoid environmentally intensive AI work in data centers. Does Foundation Models ever use Private Compute Cloud or any other kind of cloud-based AI system? I'd like to be able to assure my users that the LLM usage is relatively environmentally friendly. It would be great to be able to cite a specific Apple page explaining that Foundation Models work is always done locally. If there's any chance that work can be done in the cloud, is there a way to opt out of that?

When I use the FoundationModel framework to generate long text, it will always hit an error. "Passing along Client rate limit exceeded, try again later in response to ExecuteRequest" And stop generating. eg. for the prompt "Write a long story", it will almost certainly hit that error after 17 seconds of generation. do{ let session = LanguageModelSession() let prompt: String = "Write a long story" let response = try await session.respond(to: prompt) }catch{} If possible, I want to know how to prevent that error or at least how to handle it.

About the Core ML model encryption mention in:https://developer.apple.com/documentation/coreml/encrypting-a-model-in-your-app Ｗhen I encrypted the model, if the machine is M chip, the model will load perfectly. One the other hand, when I test the executable on an Intel chip macbook, there will be an error: Error Domain=com.apple.CoreML Code=9 "Operation not supported on this platform." UserInfo={NSLocalizedDescription=Operation not supported on this platform.} Intel test machine is 2019 macbook air with CPU: Intel i5-8210Y, OS: 14.7.6 23H626, With Apple T2 Security Chip. The encrypted model do load on M2 and M4 macbook air. If the model is NOT encrypted, it will also load on the Intel test machine. I did not find in Core ML document that suggest if the encryption/decryption support Intel chips. May I check if the decryption indeed does NOT support Intel chip?

Is it possible to train an Adaptor for the Foundation Models to produce Generable output? If so what would the response part of the training data need to look like? Presumably, under the hood, the model is outputting JSON (or some other similar structure) that can be decoded to a Generable type. Would the response part of the training data for an Adaptor need to be in that structured format?

I'm experimenting with Foundation Models and I'm trying to understand how to define a Tool whose input argument is defined at runtime. Specifically, I want a Tool that takes a single String parameter that can only take certain values defined at runtime. I think my question is basically the same as this one: https://developer.apple.com/forums/thread/793471 However, the answer provided by the engineer doesn't actually demonstrate how to create the GenerationSchema. Trying to piece things together from the documentation that the engineer linked to, I came up with this: let citiesDefinedAtRuntime = ["London", "New York", "Paris"] let citySchema = DynamicGenerationSchema( name: "CityList", properties: [ DynamicGenerationSchema.Property( name: "city", schema: DynamicGenerationSchema( name: "city", anyOf: citiesDefinedAtRuntime ) ) ] ) let generationSchema = try GenerationSchema(root: citySchema, dependencies: []) let tools = [CityInfo(parameters: generationSchema)] let session = LanguageModelSession(tools: tools, instructions: "...") With the CityInfo Tool defined like this: struct CityInfo: Tool { let name: String = "getCityInfo" let description: String = "Get information about a city." let parameters: GenerationSchema func call(arguments: GeneratedContent) throws -> String { let cityName = try arguments.value(String.self, forProperty: "city") print("Requested info about \(cityName)") let cityInfo = getCityInfo(for: cityName) return cityInfo } func getCityInfo(for city: String) -> String { // some backend that provides the info } } This compiles and usually seems to work. However, sometimes the model will try to request info about a city that is not in citiesDefinedAtRuntime. For example, if I prompt the model with "I want to travel to Tokyo in Japan, can you tell me about this city?", the model will try to request info about Tokyo, even though this is not in the citiesDefinedAtRuntime array. My understanding is that this should not be possible – constrained generation should only allow the LLM to generate an input argument from the list of cities defined in the schema. Am I missing something here or overcomplicating things? What's the correct way to make sure the LLM can only call a Tool with an input parameter from a set of possible values defined at runtime? Many thanks!

I have a fairly basic prompt I've created that parses a list of locations out of a string. I've then created a tool, which for these locations, finds their latitude/longitude on a map and populates that in the response. However, I cannot get the language model session to see/use my tool. I have code like this passing the tool to my prompt: class Parser { func populate(locations: String, latitude: Double, longitude: Double) async { let findLatLonTool = FindLatLonTool(latitude: latitude, longitude: longitude) let session = LanguageModelSession(tools: [findLatLonTool]) { """ A prompt that populates a model with a list of locations. """ """ Use the findLatLon tool to populate the latitude and longitude for the name of each location. """ } let stream = session.streamResponse(to: "Parse these locations: \(locations)", generating: ParsedLocations.self) let locationsModel = LocationsModels(); do { for try await partialParsedLocations in stream { locationsModel.parsedLocations = partialParsedLocations.content } } catch { print("Error parsing") } } } And then the tool that looks something like this: import Foundation import FoundationModels import MapKit struct FindLatLonTool: Tool { typealias Output = GeneratedContent let name = "findLatLon" let description = "Find the latitude / longitude of a location for a place name." let latitude: Double let longitude: Double @Generable struct Arguments { @Guide(description: "This is the location name to look up.") let locationName: String } func call(arguments: Arguments) async throws -> GeneratedContent { let request = MKLocalSearch.Request() request.naturalLanguageQuery = arguments.locationName request.region = MKCoordinateRegion( center: CLLocationCoordinate2D(latitude: latitude, longitude: longitude), latitudinalMeters: 1_000_000, longitudinalMeters: 1_000_000 ) let search = MKLocalSearch(request: request) let coordinate = try await search.start().mapItems.first?.location.coordinate if let coordinate = coordinate { return GeneratedContent( LatLonModel(latitude: coordinate.latitude, longitude: coordinate.longitude) ) } return GeneratedContent("Location was not found - no latitude / longitude is available.") } } But trying a bunch of different prompts has not triggered the tool - instead, what appear to be totally random locations are filled in my resulting model and at no point does a breakpoint hit my tool code. Has anybody successfully gotten a tool to be called?

Overview I'm experiencing a critical issue where TensorFlow-metal and PyArrow seem to be incompatible when installed together in the same environment. Whenever both packages are present, TensorFlow crashes and the kernel dies during execution. Environment Details Environment Details macOS Version: 15.3.2 Mac Model: MacBook Pro Max M3 Python Version: 3.11 TensorFlow Version: 2.19 PyArrow Version: 19.0.0 Issue Description: When both TensorFlow-metal and PyArrow are installed in the same Python environment, any attempt to use TensorFlow results in immediate kernel crashes. The issue appears to be a compatibility problem between these two packages rather than a problem with either package individually. Steps to Reproduce Create a new Python environment: conda create -n tf-metal python=3.11 Install TensorFlow-metal: pip install tensorflow tensorflow-metal Install PyArrow: pip install pyarrow Run the following minimal example: # Create a simple model model = tf.keras.Sequential([ tf.keras.layers.Input(shape=(2,)), tf.keras.layers.Dense(1) ]) model.compile(optimizer='adam', loss='mse') model.summary() # This works fine # Generate some dummy data X = np.random.random((100, 2)) y = np.random.random((100, 1)) # The crash happens exactly at this line model.fit(X, y, epochs=5, batch_size=32) # CRASH: Kernel dies here Result: Kernel crashes with no error message What I've Tried Reinstalling both packages in different orders Using different versions of both packages Creating isolated environments Checking system logs for additional error information The only workaround I've found is to use separate environments for each package, which isn't practical for my workflow as I need both libraries for my data processing and machine learning pipeline. Questions Has anyone else encountered this specific compatibility issue? Are there known workarounds that allow both packages to coexist? Is this a known issue that's being addressed in upcoming releases? Any insights, suggestions, or assistance would be greatly appreciated. I'm happy to provide any additional information that might help diagnose this problem. Thank you in advance for your help! Thank you in advance for your help!

I found what might be a bug with enabling Apple Intelligence when switching languages. When my iPhone's language is set to Catalan, the Apple Intelligence is disabled because it is not available for that language. Switching to Spanish doesn't activate it, and it still shows the same message of being unavailable, this time saying not available in Spanish (which is not true). However, it is enabled when the phone is rebooted. Once at this point, the bug becomes even weirder. Having the iPhone language set to Spanish and with Apple Intelligence on, I switch the language to Catalan, and the feature remains enabled. After I ask a query in Catalan, it surprisingly understands it and works, but then it gets disabled. Apart from that, as user feedback, I would love to activate Apple Intelligence in an available language other than my device's language. That's how I always used Siri (iPhone in Catalan, Siri in Spanish). Thanks!

I am trying to benchmark and see if the Qwen3 1.7B model can run in an iPhone SE 3 [4 GB RAM]. My core problem is - Even with weight quantization the SE 3 is not able to load into memory. What I've tried: I am converting a Torch model to the Core ML format using coremltools. I have tried the following combinations of quantization and context length 8 bit + 1024 8 bit + 2048 4 bit + 1024 4 bit + 2048 All the above quantizations are done with dynamic shape with the default being [1,1] in the hope that the whole context length does not get allocated in memory The 4-bit model is approximately 865MB on disk The 8-bit model is approximately 1.7 GB on disk During load: With the int4 quantization the memory spikes during intitial load a lot. Could this be because many operations are converted to int8 or fp16 as core ML does not perform operations natively on int4? With int8 on the profiler the memory does not go above 2 GB (only 900 MB) but it is still not able to load as it shows the following error. 2GB is the limit where jetsam kills the app for the iPhone SE 3 E5RT: Error(s) occurred compiling MIL to BNNS graph: [CreateBnnsGraphProgramFromMIL]: BNNS Graph Compile: failed to preallocate file with error: No space left on device for path: /var/mobile/Containers/Data/Application/ 5B8BB7D2-06A6-4BAE-A042-407B6D805E7C/Library/Caches /com.tss.qwen3-coreml/ com.apple.e5rt.e5bundlecache/ 23A341/<long key>.tmp.12586_4362093968.bundle/ H14.bundle/main/main_bnns/bnns_program.bnnsir Some online sources have suggested activation quantization but I am unsure if that will have any impact on loading [as the spike is during load and not inference] The model spec also suggests that there is no dequantization happening (for e.g from 4 bit -> fp16) So I had couple of queries: Has anyone faced similar issues? What could be the reasons for the temporary memory spike during LOAD What are approaches that can be adopted to deal with this issue? Any help would be greatly appreciated. Thank you.

Seeing this error from time to time: Context(debugDescription: "Content contains 4089 tokens, which exceeds the maximum allowed context size of 4096.", underlyingErrors: []) Of course, 4089 is less than 4096 so what is this telling me and how do I work around it? Is the limit actually lower than 4096?

As we described on the title, the model that I have built completely works on iPhone 15 / A16 Bionic, on the other hand it does not run on iPhone 16 / A18 chip with the following error message. E5RT encountered an STL exception. msg = MILCompilerForANE error: failed to compile ANE model using ANEF. Error=_ANECompiler : ANECCompile() FAILED. E5RT: MILCompilerForANE error: failed to compile ANE model using ANEF. Error=_ANECompiler : ANECCompile() FAILED (11) It consumes 1.5 ~ 1.6 GB RAM on the loading the model, then the consumption is decreased to less than 100MB on the both of iPhone 15 and 16. After that, only on iPhone 16, the above error is shown on the Xcode log, the memory consumption is surged to 5 to 6GB, and the system kills the app. It works well only on iPhone 15. This model is built with the Core ML tools. Until now, I have tried the target iOS 16 to 18 and the compute units of CPU_AND_NE and ALL. But any ways have not solved this issue. Eventually, what kindof fix should I do? minimum_deployment_target = ct.target.iOS18 compute_units = ct.ComputeUnit.ALL compute_precision = ct.precision.FLOAT16

I have built a MAC-OS machine intelligence application that uses Apple Intelligence. A part of the application is to preprocess text. For longer text content I have implemented chunking to get around the token limit. However the application performance is now limited by the fact that Apple Intelligence is sequential in operation. This has a large impact on the application performance. Is there any approach to operate Apple Intelligence in a parallel mode or even a streaming interface. As Apple Intelligence has Private Cloud Services I was hoping to be able to send multiple chunks in parallel as that would significantly improve performance. Any suggestions would be welcome. This could also be considered a request for a future enhancement.

I have a question. In China, long pressing a picture in the album can segment the target. Is this model a local model? Is there any information? Can developers use it?

Hello All, I’m working on a computer-vision–heavy iOS application that uses the camera, LiDAR depth maps, and semantic segmentation to reason about the environment (object identification, localization and measurement - not just visualization). Current architecture I initially built the image pipeline around CIImage as a unifying abstraction. It seemed like a good idea because: CIImage integrates cleanly with Vision, ARKit, AVFoundation, Metal, Core Graphics, etc. It provides a rich set of out-of-the-box transforms and filters. It is immutable and thread-safe, which significantly simplified concurrency in a multi-queue pipeline. The LiDAR depth maps, semantic segmentation masks, etc. were treated as CIImages, with conversion to CVPixelBuffer or MTLTexture only at the edges when required. Problem I’ve run into cases where Core Image transformations do not preserve numeric fidelity for non-visual data. Example: Rendering a CIImage-backed segmentation mask into a larger CVPixelBuffer can cause label values to change in predictable but incorrect ways. This occurs even when: using nearest-neighbor sampling disabling color management (workingColorSpace / outputColorSpace = NSNull) applying identity or simple affine transforms I’ve confirmed via controlled tests that: Metal → CVPixelBuffer paths preserve values correctly CIImage → CVPixelBuffer paths can introduce value changes when resampling or expanding the render target This makes CIImage unsafe as a source of numeric truth for segmentation masks and depth-based logic, even though it works well for visualization, and I should have realized this much sooner. Direction I’m considering I’m now considering refactoring toward more intent-based abstractions instead of a single image type, for example: Visual images: CIImage (camera frames, overlays, debugging, UI) Scalar fields: depth / confidence maps backed by CVPixelBuffer + Metal Label maps: segmentation masks backed by integer-preserving buffers (no interpolation, no transforms) In this model, CIImage would still be used extensively — but primarily for visualization and perceptual processing, not as the container for numerically sensitive data. Thread safety concern One of the original advantages of CIImage was that it is thread-safe by design, and that was my biggest incentive. For CVPixelBuffer / MTLTexture–backed data, I’m considering enforcing thread safety explicitly via: Swift Concurrency (actor-owned data, explicit ownership) Questions For those may have experience with CV / AR / imaging-heavy iOS apps, I was hoping to know the following: Is this separation of image intent (visual vs numeric vs categorical) a reasonable architectural direction? Do you generally keep CIImage at the heart of your pipeline, or push it to the edges (visualization only)? How do you manage thread safety and ownership when working heavily with CVPixelBuffer and Metal? Using actor-based abstractions, GCD, or adhoc? Are there any best practices or gotchas around using Core Image with depth maps or segmentation masks that I should be aware of? I’d really appreciate any guidance or experience-based advice. I suspect I’ve hit a boundary of Core Image’s design, and I’m trying to refactor in a way that doesn't involve too much immediate tech debt, remains robust and maintainable long-term. Thank you in advance!

I am trying to test FoundationModels in a Swift Playground in Xcode 26.2, macOS 26.3, and am running into an issue. The following simple code generates an error: import FoundationModels @Generable struct Specifications { @Guide(description: "Search for color") var color: String } I see the following error message in the console: error: AIPlayground.playground:4:8: external macro implementation type 'FoundationModelsMacros.GenerableMacro' could not be found for macro 'Generable(description:)'; plugin for module 'FoundationModelsMacros' not found The Xcode editor does not appear to recognize the @Generable or @Guide macros, despite importing FoundationModels. What step/setting am I missing?

A foundation models bug I keep running into when in the preview phase of the testing. The error never seems to occur or break the app when I am testing on the simulator or on a device but sometimes I am running into this error when in a longer session while being in preview. The error breaks the preview and crashes it and the waring on it is labeled as : "Assert in LanguageModelFeedback.swift" This is something I keep running into, where I have been using foundation models for my project