At WWDC25 we launched a new type of Lab event for the developer community - Group Labs. A Group Lab is a panel Q&A designed for a large audience of developers. Group Labs are a unique opportunity for the community to submit questions directly to a panel of Apple engineers and designers. Here are the highlights from the WWDC25 Group Lab for Machine Learning and AI Frameworks. What are you most excited about in the Foundation Models framework? The Foundation Models framework provides access to an on-device Large Language Model (LLM), enabling entirely on-device processing for intelligent features. This allows you to build features such as personalized search suggestions and dynamic NPC generation in games. The combination of guided generation and streaming capabilities is particularly exciting for creating delightful animations and features with reliable output. The seamless integration with SwiftUI and the new design material Liquid Glass is also a major advantage. When should I still bring my own LLM via CoreML? It's generally recommended to first explore Apple's built-in system models and APIs, including the Foundation Models framework, as they are highly optimized for Apple devices and cover a wide range of use cases. However, Core ML is still valuable if you need more control or choice over the specific model being deployed, such as customizing existing system models or augmenting prompts. Core ML provides the tools to get these models on-device, but you are responsible for model distribution and updates. Should I migrate PyTorch code to MLX? MLX is an open-source, general-purpose machine learning framework designed for Apple Silicon from the ground up. It offers a familiar API, similar to PyTorch, and supports C, C++, Python, and Swift. MLX emphasizes unified memory, a key feature of Apple Silicon hardware, which can improve performance. It's recommended to try MLX and see if its programming model and features better suit your application's needs. MLX shines when working with state-of-the-art, larger models. Can I test Foundation Models in Xcode simulator or device? Yes, you can use the Xcode simulator to test Foundation Models use cases. However, your Mac must be running macOS Tahoe. You can test on a physical iPhone running iOS 18 by connecting it to your Mac and running Playgrounds or live previews directly on the device. Which on-device models will be supported? any open source models? The Foundation Models framework currently supports Apple's first-party models only. This allows for platform-wide optimizations, improving battery life and reducing latency. While Core ML can be used to integrate open-source models, it's generally recommended to first explore the built-in system models and APIs provided by Apple, including those in the Vision, Natural Language, and Speech frameworks, as they are highly optimized for Apple devices. For frontier models, MLX can run very large models. How often will the Foundational Model be updated? How do we test for stability when the model is updated? The Foundation Model will be updated in sync with operating system updates. You can test your app against new model versions during the beta period by downloading the beta OS and running your app. It is highly recommended to create an "eval set" of golden prompts and responses to evaluate the performance of your features as the model changes or as you tweak your prompts. Report any unsatisfactory or satisfactory cases using Feedback Assistant. Which on-device model/API can I use to extract text data from images such as: nutrition labels, ingredient lists, cashier receipts, etc? Thank you. The Vision framework offers the RecognizeDocumentRequest which is specifically designed for these use cases. It not only recognizes text in images but also provides the structure of the document, such as rows in a receipt or the layout of a nutrition label. It can also identify data like phone numbers, addresses, and prices. What is the context window for the model? What are max tokens in and max tokens out? The context window for the Foundation Model is 4,096 tokens. The split between input and output tokens is flexible. For example, if you input 4,000 tokens, you'll have 96 tokens remaining for the output. The API takes in text, converting it to tokens under the hood. When estimating token count, a good rule of thumb is 3-4 characters per token for languages like English, and 1 character per token for languages like Japanese or Chinese. Handle potential errors gracefully by asking for shorter prompts or starting a new session if the token limit is exceeded. Is there a rate limit for Foundation Models API that is limited by power or temperature condition on the iPhone? Yes, there are rate limits, particularly when your app is in the background. A budget is allocated for background app usage, but exceeding it will result in rate-limiting errors. In the foreground, there is no rate limit unless the device is under heavy load (e.g., camera open, game mode). The system dynamically balances performance, battery life, and thermal conditions, which can affect the token throughput. Use appropriate quality of service settings for your tasks (e.g., background priority for background work) to help the system manage resources effectively. Do the foundation models support languages other than English? Yes, the on-device Foundation Model is multilingual and supports all languages supported by Apple Intelligence. To get the model to output in a specific language, prompt it with instructions indicating the user's preferred language using the locale API (e.g., "The user's preferred language is en-US"). Putting the instructions in English, but then putting the user prompt in the desired output language is a recommended practice. Are larger server-based models available through Foundation Models? No, the Foundation Models API currently only provides access to the on-device Large Language Model at the core of Apple Intelligence. It does not support server-side models. On-device models are preferred for privacy and for performance reasons. Is it possible to run Retrieval-Augmented Generation (RAG) using the Foundation Models framework? Yes, it is possible to run RAG on-device, but the Foundation Models framework does not include a built-in embedding model. You'll need to use a separate database to store vectors and implement nearest neighbor or cosine distance searches. The Natural Language framework offers simple word and sentence embeddings that can be used. Consider using a combination of Foundation Models and Core ML, using Core ML for your embedding model.

Hi! I noticed that on my father's M1 Max MacBook Pro (64gb ram) there's an option for style transfer which I don't see on my M1 MacBook Air (16gb ram). I am running macOS Tahoe and he is running macOS Sequoia.

Hi i'm curently crating a model to identify car plates (object detection) i use asitop to monitor my macbook pro and i see that only the cpu is used for the training and i wanted to know why

With the release of the newest version of tahoe and MLX supporting RDMA. Is there a documentation link to how to utilizes the libdrma dylib as well as what functions are available? I am currently assuming it mostly follows the standard linux infiniband library but I would like the apple specific details.

It seems to be that Swift has more APIs implemented than the C++ interface (especially APIs found in the MLXNN and MLXOptimize folders). Is there any intention to implement more APIs for neural networks and training them in the future?

Hello everyone, I’m currently working with the Message Filtering Extension and would really appreciate some clarification around its performance and operational constraints. While the extension is extremely powerful and useful, I’ve found that some important details are either unclear or not well covered in the available documentation. There are two main areas I’m trying to understand better: Machine learning model constraints within the extension In our case, we already have an existing ML model that classifies messages (and are not dependant on Apple's built-in models). We’re evaluating whether and how it can be used inside the extension. Specifically, I’m trying to understand: Are there documented limits on the size of an ML model (e.g., maximum bundle size or model file size in MB)? What are the memory constraints for a model once loaded into memory by the extension? Under what conditions would the system terminate or “kick out” the extension due to memory or performance pressure? Message processing timeouts and execution constraints What is the timeout for processing a single received message? At what point will the OS stop waiting for the extension’s response and allow the message by default (for example, if the extension does not respond in time)? Any guidance, official references, or practical experience from Apple engineers or other developers would be greatly appreciated. Thanks in advance for your help,

The developer tutorial for visual intelligence indicates that the method to detect and handle taps on a displayed entity from the Search section is via an "OpenIntent" associated with your entity. However, running this intent executes code from within my app. If I have the perform() method display UI, it always displays UI from within my app. I noticed that the Google app's integration to visual intelligence has a different behavior-- tapping on an entity does not take you to the Google app -- instead, a Webview is presented sheet-style WITHIN the Visual Intelligence environment (see below) How is that accomplished?

I’m considering creating an ILMessageFilterExtension using a mini LLM/SLM to detect fraud and I’ve read it has strict memory limits yet I can’t find it in the documentation. What’s the set limit or any other constraints impacting the feasibility of running 100-500mb model?

: Hello, I’m seeking clarification on whether Apple provides any framework or API that enables deep integration between Siri and advanced AI assistants (such as ChatGPT), including system-level functions like voice interaction, navigation, cross-platform syncing, and operational access similar to Siri’s own capabilities. If no such option exists today, I would appreciate guidance on the recommended path or approved third-party solutions for building a unified, voice-first experience across Apple’s ecosystem. Thank you for your time and insight.

We built an open-source macOS menu bar app that turns speech into text and pastes it into the active app — using SpeechAnalyzer for on-device transcription, ScreenCaptureKit + Vision for screen-aware context, and FluidAudio for speaker diarization in meeting mode. Here's what we learned shipping it on macOS 26. GitHub: github.com/Marvinngg/ambient-voice Architecture The app has two modes: hotkey dictation (press to talk, release to inject) and meeting recording (continuous transcription with a floating panel). Dictation Mode Audio capture uses AVCaptureSession (more on why below). The captured audio feeds into SpeechAnalyzer via an AsyncStream: let transcriber = SpeechTranscriber( locale: locale, transcriptionOptions: [], reportingOptions: [.volatileResults, .alternativeTranscriptions], attributeOptions: [.audioTimeRange, .transcriptionConfidence] ) let analyzer = SpeechAnalyzer(modules: [transcriber]) let (inputSequence, inputBuilder) = AsyncStream.makeStream() try await analyzer.start(inputSequence: inputSequence) While recording, we capture a screenshot of the focused window using ScreenCaptureKit, run Vision OCR (VNRecognizeTextRequest), extract keywords, and inject them into SpeechAnalyzer as contextual bias: let context = AnalysisContext() context.contextualStrings[.general] = ocrKeywords try await analyzer.setContext(context) This improves accuracy for technical terms and proper nouns visible on screen. If your screen shows "SpeechAnalyzer", saying it out loud is more likely to be transcribed correctly. After transcription, an optional L2 step sends the text through a local LLM (ollama) for spoken-to-written cleanup, then CGEvent simulates Cmd+V to paste into the active app. Meeting Mode Meeting mode forks the same audio stream to two consumers: SpeechAnalyzer — real-time streaming transcription, displayed in a floating NSPanel FluidAudio buffer — accumulates 16kHz Float32 mono samples for batch speaker diarization after recording stops When the user ends the meeting, FluidAudio's performCompleteDiarization() runs on the accumulated audio. We align transcription segments with speaker segments using audioTimeRange overlap matching — each transcription segment gets assigned the speaker ID with the most time overlap. Results export to Markdown. Pitfalls We Hit on macOS 26 1. AVAudioEngine installTap doesn't fire with Bluetooth devices We started with AVAudioEngine.inputNode.installTap() for audio capture. It worked fine with built-in mics but the tap callback never fired with Bluetooth devices (tested with vivo TWS 4 Hi-Fi). Fix: switched to AVCaptureSession. The delegate callback captureOutput(_:didOutput:from:) fires reliably regardless of audio device. The tradeoff is you get CMSampleBuffer instead of AVAudioPCMBuffer, so you need a conversion step. 2. NSEvent addGlobalMonitorForEvents crashes Our global hotkey listener used NSEvent.addGlobalMonitorForEvents. On macOS 26, this crashes with a Bus error inside GlobalObserverHandler — appears to be a Swift actor runtime issue. Fix: switched to CGEventTap. Works reliably, but the callback runs on a CFRunLoop context, which Swift doesn't recognize as MainActor. 3. CGEventTap callbacks aren't on MainActor If your CGEventTap callback touches any @MainActor state, you'll get concurrency violations. The callback runs on whatever thread owns the CFRunLoop. Fix: bridge with DispatchQueue.main.async {} inside the tap callback before touching any MainActor state. 4. CGPreflightScreenCaptureAccess doesn't request permission We used CGPreflightScreenCaptureAccess() as a guard before calling ScreenCaptureKit. If it returned false, we'd bail out. The problem: this function only checks — it never triggers macOS to add your app to the Screen Recording permission list. Chicken-and-egg: you can't get permission because you never ask for it. Fix: call CGRequestScreenCaptureAccess() at app startup. This adds your app to System Settings → Screen Recording. Then let ScreenCaptureKit calls proceed without the preflight guard — SCShareableContent will also trigger the permission prompt on first use. 5. Ad-hoc signing breaks TCC permissions on every rebuild During development, codesign --sign - (ad-hoc) generates a different code directory hash on every build. macOS TCC tracks permissions by this hash, so every rebuild = new app identity = all permissions reset. Fix: sign with a stable certificate. If you have an Apple Development certificate, use that. The TeamIdentifier stays constant across rebuilds, so TCC permissions persist. We also discovered that launching via open WE.app (LaunchServices) instead of directly executing the binary is required — otherwise macOS attributes TCC permissions to Terminal, not your app. Benchmarks We ran end-to-end benchmarks on public datasets (Mac Mini M4 16GB, macOS 26): Transcription (SpeechAnalyzer, AliMeeting Chinese): • Near-field CER 34% (excluding outliers ~25%) • Far-field CER 40% (single channel, no beamforming, >30% overlap) • Processing speed 74-89x real-time Speaker diarization (FluidAudio offline): • AMI English 16 meetings: avg DER 23.2% (collar=0.25s, ignoreOverlap=True) • AliMeeting Chinese 8 meetings: DER 48.5% (including overlap regions) • Memory: RSS ~500MB, peak 730-930MB Full evaluation methodology, scripts, and raw results are in the repo. Open Source The project is MIT licensed: github.com/Marvinngg/ambient-voice It includes the macOS client (Swift 6.2, SPM), server-side distillation/training scripts (Python), and a complete evaluation framework with reproducible benchmarks. Feedback and contributions welcome.

After a few attempts I end up at the same spot as categorized here by ChatGPT - I get why it feels that way, Rob — honestly, you’ve walked right into the worst possible moment in Apple’s Core ML toolchain evolution. What you’re running into isn’t your fault; Apple really did rip out the ONNX front-end without offering a clean replacement yet. So every tutorial, every script, and even coremltools itself contradicts what’s actually possible right now. Let’s be brutally clear about the state of things (late 2025): Goal Supported by Apple today Practical workaround Convert ONNX → Core ML using coremltools 9 No. The ONNX importer was removed. Use the legacy coremltools 6.x + onnx-coreml stack in a separate Python 3.9/3.10 env. Convert PyTorch → Core ML directly Yes, but only for simple or traced models. T5 still fails because of dynamic control flow. None reliable without rewriting the model. Run ONNX models directly on iOS Partially (via MLC LLM / Metal Performance Shaders). Works, but that’s an entirely different runtime. So: you’re not crazy, and this isn’t wasted effort—you’ve actually proven that your model exports cleanly and that the only missing piece is Apple’s converter. The ONNX file you built is solid. You just need something that can translate it to Core ML. If your goal is simply “get it running on-device”, you still have two viable moves that aren’t rabbit holes: 1. Use MLC LLM / Metal runtime Skip Core ML altogether. Apple’s mlc-llm (and the “MLCChat” pipeline) can execute your ONNX or quantized model directly on the GPU. You keep full fidelity and avoid all these broken converters. 2. Use the legacy Core ML 6.x toolchain in a sandboxed Python 3.10 VM It actually works. You don’t have to downgrade your whole system—just run that specific environment once to emit the .mlpackage, then never touch it again. Nothing else will silently fix itself until Apple re-adds ONNX ingestion in Core ML 10 or a new converter.

I used Yolo5-11 and while performing great detecting balls lets say 5-10ft away in 1920 resolution and even in 640 it really is taking toll on my app performance. When I use Create ML it outputs all in 415x which is probably the reason why it does not detect objects from far. What can I do to preserve some energy ? My model is used with about 1K pictures 200 each test and validate, and from close up and far.

I have an app that uses a couple of mlmodels (word tagger and gazetteer) and I’m trying to encrypt them before publishing. The models are part of a package. I understand that Xcode can’t automatically handle the encryption for a model in a package the way it can within a traditional app structure. Given that, I’ve generated the Apple MLModel encryption key from Xcode and am encrypting via the command line with: xcrun coremlcompiler compile Gazetteer.mlmodel GazetteerENC.mlmodelc --encrypt Gazetteerkey.mlmodelkey In the package manifest, I’ve listed the encrypted models as .copy resources for my target and have verified the URL to that file is good. When I try to load the encrypted .mlmodelc file (on a physical device) with the line:
 gazetteer = try NLGazetteer(contentsOf: gazetteerURL!) I get the error: Failed to open file: /…/Scanner.bundle/GazetteerENC.mlmodelc/coremldata.bin. It is not a valid .mlmodelc file. So my questions are: Does the NLGazetteer class support encrypted MLModel files? Given that my models are in a package, do I have the right general approach? Thanks for any help or thoughts.

I've built a model using Create ML, but I can't make it, for the love of God, updatable. I can't find any checkbox or anything related. It's an Activity Classifier, if it matters. I want to continue training it on-device using MLUpdateTask, but the model, as exported from Create ML, fails with error: Domain=com.apple.CoreML Code=6 "Failed to unarchive update parameters. Model should be re-compiled." UserInfo={NSLocalizedDescription=Failed to unarchive update parameters. Model should be re-compiled.}

I get the following error when running this command in a Jupyter notebook: v = tf.Variable(initial_value=tf.random.normal(shape=(3, 1))) v[0, 0].assign(3.) Environment: python == 3.11.14 tensorflow==2.19.1 tensorflow-metal==1.2.0 { "name": "InvalidArgumentError", "message": "Cannot assign a device for operation ResourceStridedSliceAssign: Could not satisfy explicit device specification '/job:localhost/replica:0/task:0/device:GPU:0' because no supported kernel for GPU devices is available.\nColocation Debug Info:\nColocation group had the following types and supported devices: \nRoot Member(assigned_device_name_index_=1 requested_device_name_='/job:localhost/replica:0/task:0/device:GPU:0' assigned_device_name_='/job:localhost/replica:0/task:0/device:GPU:0' resource_device_name_='/job:localhost/replica:0/task:0/device:GPU:0' supported_device_types_=[CPU] possible_devices_=[]\nResourceStridedSliceAssign: CPU \n_Arg: GPU CPU \n\nColocation members, user-requested devices, and framework assigned devices, if any:\n ref (_Arg) framework assigned device=/job:localhost/replica:0/task:0/device:GPU:0\n ResourceStridedSliceAssign (ResourceStridedSliceAssign) /job:localhost/replica:0/task:0/device:GPU:0\n\nOp: ResourceStridedSliceAssign\n [...] [[{{node ResourceStridedSliceAssign}}]] [Op:ResourceStridedSliceAssign] name: strided_slice/_assign" } It seems like the ResourceStridedSliceAssign operation is not implemented for the GPU

Can't able to run the Create ML for training and I upgraded to MacOS 26.3 beta and I have tried older and newer

Hi all, I'm trying to find out if/when we can expect mxfp8/mxfp4 support on Apple Silicon. I've noticed that mlx now has casting data types, but all computation is still done in bf16. Would be great to reduce power consumption with support for these lower precision data types since edge inference is already typically done at a lower precision! Thanks in advance.

Hi Apple Engineers, I am experiencing a potential memory management bug with CoreML on M1 Mac (32GB Unified Memory). When processing long video files (approx. 12,000 frames) using a CoreML execution provider, the system often completes the 'Analysing' phase but fails to transition into 'Processing'. It simply exits silently or hits an import error (scipy). However, if I split the same task into small 20-frame segments, it works perfectly at high speeds (~40 FPS). This suggests the hardware is capable, but there is an issue with memory fragmentation or resource cleanup during long-running CoreML sessions. Is there a way to force a VRAM/Unified Memory flush via CLI, or is this a known limitation for large frame indexing?

Hello everyone, I’m looking for guidance regarding my app review timeline, as things seem unusually delayed compared to previous submissions. My iOS app was rejected on November 19th due to AI-related policy questions. I immediately responded to the reviewer with detailed explanations covering: Model used (Gemini Flash 2.0 / 2.5 Lite) How the AI only generates neutral, non-directive reflective questions How the system prevents any diagnosis, therapy-like behavior or recommendations Crisis-handling limitations Safety safeguards at generation and UI level Internal red-team testing and results Data retention, privacy, and non-use of data for model training After sending the requested information, I resubmitted the build on November 19th at 14:40. Since then: November 20th (7:30) → Status changed to In Review. November 21st, 22nd, 23rd, 24th, 25th → No movement, still In Review. My open case on App Store Connect is still pending without updates. Because of the previous rejection, I expected a short delay, but this is now 5 days total and 3 business days with no progress, which feels longer than usual for my past submissions. I’m not sure whether: My app is in a secondary review queue due to the AI-related rejection, The reviewer is waiting for internal clarification, Or if something is stuck and needs to be escalated. I don’t want to resubmit a new build unless necessary, since that would restart the queue. Could someone from the community (or Apple, if possible) confirm whether this waiting time is normal after an AI-policy rejection? And is there anything I should do besides waiting — for example, contacting Developer Support again or requesting a follow-up? Thank you very much for your help. I appreciate any insight from others who have experienced similar delays.

I watched this year WWDC25 "Read Documents using the Vision framework". At the end of video there is mention of new DetectHandPoseRequest model for hand pose detection in Vision API. I looked Apple documentation and I don't see new revision. Moreover probably typo in video because there is only DetectHumanPoseRequst (swift based) and VNDetectHumanHandPoseRequest (obj-c based) (notice lack of Human prefix in WWDC video) First one have revision only added in iOS 18+: https://developer.apple.com/documentation/vision/detecthumanhandposerequest/revision-swift.enum/revision1 Second one have revision only added in iOS14+: https://developer.apple.com/documentation/vision/vndetecthumanhandposerequestrevision1 I don't see any new revision targeting iOS26+

Hi everyone, I believe I’ve encountered a potential bug or a hardware alignment limitation in the Core ML Framework / ANE Runtime specifically affecting the new Stateful API (introduced in iOS 18/macOS 15). The Issue: A Stateful mlprogram fails to run on the Apple Neural Engine (ANE) if the state tensor dimensions (specifically the width) are not a multiple of 32. The model works perfectly on CPU and GPU, but fails on ANE both during runtime and when generating a Performance Report in Xcode. Error Message in Xcode UI: "There was an error creating the performance report Unable to compute the prediction using ML Program. It can be an invalid input data or broken/unsupported model." Observations: Case A (Fails): State shape = (1, 3, 480, 270). Prediction fails on ANE. Case B (Success): State shape = (1, 3, 480, 256). Prediction succeeds on ANE. This suggests an internal memory alignment or tiling issue within the ANE driver when handling Stateful buffers that don't meet the 32-pixel/element alignment. Reproduction Code (PyTorch + coremltools): import torch.nn as nn import coremltools as ct import numpy as np class RNN_Stateful(nn.Module): def __init__(self, hidden_shape): super(RNN_Stateful, self).__init__() # Simple conv to update state self.conv1 = nn.Conv2d(3 + hidden_shape[1], hidden_shape[1], kernel_size=3, padding=1) self.conv2 = nn.Conv2d(hidden_shape[1], 3, kernel_size=3, padding=1) self.register_buffer("hidden_state", torch.ones(hidden_shape, dtype=torch.float16)) def forward(self, imgs): self.hidden_state = self.conv1(torch.cat((imgs, self.hidden_state), dim=1)) return self.conv2(self.hidden_state) # h=480, w=255 causes ANE failure. w=256 works. b, ch, h, w = 1, 3, 480, 255 model = RNN_Stateful((b, ch, h, w)).eval() traced_model = torch.jit.trace(model, torch.randn(b, 3, h, w)) mlmodel = ct.convert( traced_model, inputs=[ct.TensorType(name="input_image", shape=(b, 3, h, w), dtype=np.float16)], outputs=[ct.TensorType(name="output", dtype=np.float16)], states=[ct.StateType(wrapped_type=ct.TensorType(shape=(b, ch, h, w), dtype=np.float16), name="hidden_state")], minimum_deployment_target=ct.target.iOS18, convert_to="mlprogram" ) mlmodel.save("rnn_stateful.mlpackage") Steps to see the error: Open the generated .mlpackage in Xcode 16.0+. Go to the Performance tab and run a test on a device with ANE (e.g., iPhone 15/16 or M-series Mac). The report will fail to generate with the error mentioned above. Environment: OS: macOS 15.2 Xcode: 16.3 Hardware: M4 Has anyone else encountered this 32-pixel alignment requirement for StateType tensors on ANE? Is this a known hardware constraint or a bug in the Core ML runtime? Any insights or workarounds (other than manual padding) would be appreciated.