Hi, I’m developing an app targeting iOS 26, using the new FoundationModels framework to perform on-device LLM inference. I’m currently testing memory usage. Does the memory used by FoundationModels—including model weights, KV cache, and any inference-related buffers—count toward my app’s Jetsam memory limit, or is any of it managed separately by the system? I may need to run two concurrent inferences, each with a 4096-token context window. Is this explicitly supported or allowed by FoundationModels on iOS 26? Would this significantly increase the risk of memory-based termination? Thanks in advance for any clarification. Thanks.

I am trying to use @AppIntent(schema: .system.search) to search in my app via a Siri voice command, but I want to be able to return a .result that does not open the app, yet still get the model training benefits from the schema. Very new to this, this is my first app, so I would appreciate some guidance. I haven't gotten to the voice part, I tested on Shortcuts. Do I need to do AppIntents without the schema and wait until there is a search schema that does not open the app, or should I be using a different schema? What am I missing?

How reliable is the Models, to use as a comparison, such as a cholesterol test, to inform, for example, whether it is worth it to go see a doctor? I would like to use Tool to attach the simple blood test data to the session and with this the Model can analyse and made a simple suggestion if is necessary to see a doctor etc.. ? ps.: Local model

Encountered a few times when the answer get "stuck" (I am now at beta 6). This is an example.

Hello Apple Developer Community, I'm investigating Core ML model loading behavior and noticed that even when the compiled model path remains unchanged after an APP update, the first run still triggers an "uncached load" process. This seems to impact user experience with unnecessary delays. Question: Does Core ML provide any public API to check whether a compiled model (from a specific .mlmodelc path) is already cached in the system? If such API exists, we'd like to use it for pre-loading decision logic - only perform background pre-load when the model isn't cached. Has anyone encountered similar scenarios or found official solutions? Any insights would be greatly appreciated!

Is there anywhere we can reference error codes? I'm getting this error: "The operation couldn’t be completed. (FoundationModels.LanguageModelSession.GenerationError error 4.)" and I have no idea of what it means or what to attempt to fix.

I'm building a new feature with Visual Intelligence framework. My implementation for IndexedEntity and IntentValueQuery worked as expected and I can see a list of objects in visual search result. However, my OpenIntent doesn't work. When I tap on the object, I got a message on screen "Sorry somethinf went wrong ...". and the breakpoint in perform() is never triggered. Things I've tried: I added @MainActor before perform(), this didn't change anything I set static let openAppWhenRun: Bool = true and static var supportedModes: IntentModes = [.foreground(.immediate)], still nothing I created a different intent for the see more button at the end of feed. This AppIntent with schema: .visualIntelligence.semanticContentSearch worked, perform() is executed

I'm testing Foundation Model on my iPad Pro (5th gen) iOS 26. Up until late this morning, I can no longer load the SystemLanguageModel.default. I'm not doing anything interesting, something as basic as this is only going to unavailable, specifically I get unavailable reason: modelNotReady. let model = SystemLanguageModel.default ... switch model.availability { case .available: print("LM available") case .unavailable(let reason): print("unavailable reason: ", String(describing: reason)) } I also ran the FoundationModelsTripPlanner app, same thing. It was working yesterday, I have not modified that project either. Why is the Model not ready? How do I fix this? Yes, I tried restarting both my laptop and iPad, no luck.

Hey, I've been trying to write an AI agent for OpenAI's GPT-5, but using the @Generable Tool types from the FoundationModels framework, which is super awesome btw! I'm having trouble implementing the tool calling, though. When I receive a tool call from the OpenAI api, I do the following: Find the tool in my [any Tool] array via the tool name I get from the model if let tool = tools.first(where: { $0.name == functionCall.name }) { // ... } Parse the arguments of the tool call via GeneratedContent(json:) let generatedContent = try GeneratedContent(json: functionCall.arguments) Pass the tool and arguments to a function that calls tool.call(arguments: arguments) and returns the tool's output type private func execute<T: Tool>(_ tool: T, with generatedContent: GeneratedContent) async throws -> T.Output { let arguments = try T.Arguments.init(generatedContent) return try await tool.call(arguments: arguments) } Up to this point, everything is working as expected. However, the tool's output type is any PromptRepresentable and I have no idea how to turn that into something that I can encode and send back to the model. I assumed there might be a way to turn it into a GeneratedContent but there is no fitting initializer. Am I missing something or is this not supported? Without a way to return the output to an external provider, it wouldn't really be possible to use FoundationModels Tool type I think. That would be unfortunate because it's implemented so elegantly. Thanks!

I've run into an issue with a small Foundation Models test with Generable. I'm getting a strange error message with this Generable. I was able to get simpler ones to work. Is this because the Generable is recursive with a property of [HTMLDiv]? The error message is: FoundationModels/SchemaAugmentor.swift:209: Fatal error: 'try!' expression unexpectedly raised an error: FoundationModels.GenerationSchema.SchemaError.undefinedReferences(schema: Optional("SafeResponse<HTMLDiv>"), references: ["HTMLDiv"], context: FoundationModels.GenerationSchema.SchemaError.Context(debugDescription: "Undefined types: [HTMLDiv]", underlyingErrors: [])) The code is: import FoundationModels import Playgrounds @Generable struct HTMLDiv { @Guide(description: "Optional named ID, useful for nicknames") var id: String? = nil @Guide(description: "Optional visible HTML text") var textContent: String? = nil @Guide(description: "Any child elements", .count(0...10)) var children: [HTMLDiv] = [] static var sample: HTMLDiv { HTMLDiv( id: "profileToolbar", children: [ HTMLDiv(textContent: "Log in"), HTMLDiv(textContent: "Sign up"), ] ) } } #Playground { do { let session = LanguageModelSession { "Your job is to generate simple HTML markup" "Here is an example response to the prompt: 'Make a profile toolbar':" HTMLDiv.sample } let response = try await session.respond( to: "Make a sign up form", generating: HTMLDiv.self ) print(response.content) } catch { print(error) } }

Hello, My app fully relies on the new Foundation Models. Since Foundation Models require Apple Intelligence, I want to ensure that only devices capable of running Apple Intelligence can install my app. When checking the UIRequiredDeviceCapabilities property for a suitable value, I found that iphone-performance-gaming-tier seems the closest match. Based on my research: On iPhone, this effectively limits installation to iPhone 15 Pro or later. On iPad, it ensures M1 or newer devices. This exactly matches the hardware requirements for Apple Intelligence. However, after setting iphone-performance-gaming-tier, I noticed that on iPad, Game Mode (Game Overlay) is automatically activated, and my app is treated as a game. My questions are: Is there a more appropriate UIRequiredDeviceCapabilities value that would enforce the same Apple Intelligence hardware requirements without triggering Game Mode? If not, is there another way to restrict installation to devices meeting Apple Intelligence requirements? Is there a way to prevent Game Mode from appearing for my app while still using this capability restriction? Thanks in advance for your help.

Environment: Hardware: Mac M4 OS: macOS Sequoia 15.7.4 TensorFlow-macOS Version: 2.16.2 TensorFlow-metal Version: 1.2.0 Description: When using the tensorflow-metal plug-in for GPU acceleration on M4, the ReLU activation function (both as a layer and as an activation argument) fails to correctly clip negative values to zero. The same code works correctly when forced to run on the CPU. Reproduction Script: import os import numpy as np import tensorflow as tf # weights and biases = -1 weights = [np.ones((10, 5)) * -1, np.ones(5) * -1] # input = 1 data = np.ones((1, 10)) # comment this line => GPU => get negative values # uncomment this line => CPU => no negative values # tf.config.set_visible_devices([], 'GPU') # create model model = tf.keras.Sequential([ tf.keras.layers.Input(shape=(10,)), tf.keras.layers.Dense(5, activation='relu') ]) # set weights model.layers[0].set_weights(weights) # get output output = model.predict(data) # check if negative is present print(f"min value: {output.min()}") print(f"is negative present? {np.any(output < 0)}")

I'm a bit new to the LLM stuff and with Foundation Models. My understanding is that there is a token limit of around 4K. I want to process the contents of files which may be quite large. I first tried going the Tool route but that didn't work out so I then tried manually chunking the text to keep things under the limit. It mostly works except that every now and then it'll exceed the limit. This happens even when the chunks are less than 100 characters. Instructions themselves are about 500 characters but still overall, well below 1000 characters per prompt, all told, which, in my limited understanding, should not result in 4K tokens being parsed. Any ideas on what is going on here?

When context window size exceeded, this error is not called (instead another error has shown up) to handle new session. LanguageModelSession.GenerationError.exceededContextWindowSize Or am I doing things wrong?

Hello Apple Team, Thank you for the recent Group Lab and for your continued work on advancing Xcode and developer tools. I’d like to submit a feature request: Are there any plans to introduce support for Agentic AI Mode (MCP protocol) in future versions of iOS or Xcode? As developer tools evolve toward more intelligent and context-aware environments, the integration of agentic AI capabilities could significantly enhance productivity and unlock new creative workflows. Looking forward to your consideration, and thank you again for the excellent session. Best regards

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

I’m trying to integrate Apple’s Translation framework in a Swift 6 project with Approachable Concurrency enabled. I’m following the code here: https://developer.apple.com/documentation/translation/translating-text-within-your-app#Offer-a-custom-translation And, specifically, inside the following code .translationTask(configuration) { session in do { // Use the session the task provides to translate the text. let response = try await session.translate(sourceText) // Update the view with the translated result. targetText = response.targetText } catch { // Handle any errors. } } On the try await session.translate(…) line, the compiler complains that “Sending ‘session’ risks causing data races”. Extended error message: Sending main actor-isolated 'session' to @concurrent instance method 'translate' risks causing data races between @concurrent and main actor-isolated uses I’ve downloaded Apple’s sample code (at the top of linked webpage), it compiles fine as-is on Xcode 26.4, but fails with the same error as soon as I switch the Swift Language Mode to Swift 6 in the project. How can I fix this?

Do we know what a safe max token limit is? After some iterating, I have come to believe 4096 might be the limit on device. Could you help me out by answering any of these questions: Is 4096 the correct limit? Do all devices have the same limit? Will the limit change over time or by device? The errors I get when going over the limit do not seem to say, hey you are over, so it's just by trial and error that I figure these issues out. Thanks for the fun new toys. Regards, Rob

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?

Reading all the docs(1) I was under the impression that handling this error is well managed... Until I hit it and found out that the recommended handling options hide a crucial fact: in the catch block you can not do anything?! It's too late - everything is lost, no way to recover... All the docs mislead me that I can apply the Transcript trick in the catch block until I realised, that there is nothing there !!! This article here(2) enlightened me on the handling of this problem, but I must say (and the author as well) - this is a hack! So my questions: is there really no way to handle this exception properly? if not, can we have the most important information - the count of the context exposed through the official API (at least the known ones)? https://developer.apple.com/documentation/Technotes/tn3193-managing-the-on-device-foundation-model-s-context-window#Handle-the-exceeding-context-window-size-error-elegantly https://zats.io/blog/making-the-most-of-apple-foundation-models-context-window/