Hey, When generating responses with structured output and non-streaming API, it sometimes takes 3s, sometimes 10-20s. I am firing that request subsequently while testing the app. Is this by design, or any place I can learn more about what contributes to such variation?

I am using gemini2.5-flash with SwiftUI. How can I receive a response in JSON?

Hi everyone 👋 I'd like to use coremltools to see how well a model performs on a remote device as part of a CI/CD pipeline. According to the Core ML Tools "Debugging and Performance Utilities" guide, remote devices must be in a "connected" state in order for coremltools to install the ModelRunner application. The devices in our system have a "paired" state, and I'm unable to set the them as "connected." The only way I know how to connect a device is to physically plug it in to a computer and open Xcode. I don't have physical access to the devices in the CI/CD system, and the host computer that interacts with them doesn't have Xcode installed. Here are some questions I've been looking into and would love some help answering: Has anyone managed to use the coremltools performance utilities in a similar system? Can you put a device in a "connected" state if you don't have physical access to the device and if you only have access to Xcode command line tools and not the Xcode app? Is it at all possible to install the coremltools ModelRunner application on a "paired" device, for example, by manually building the app and installing it with devicectl? Would other utilities, such as the MLModelBenchmarker work as expected if the app is installed this way? Thank you!

I've built an iOS app with a novel approach to AI safety: a deterministic, pre-inference validation layer called Newton Engine. Instead of relying on the LLM to self-moderate, Newton validates every prompt BEFORE it reaches the model. It uses shape theory and semantic analysis to detect: • Corrosive frames (self-harm language patterns) • Logical contradictions (requests that undermine themselves) • Delegation attempts (asking AI to make human decisions) • Jailbreak patterns (prompt injection, role-play escapes) • Hallucination triggers (requests for fabricated citations) The system achieves a 96% adversarial catch rate across 847 test cases, with zero false positives on benign prompts. Key technical details: • Pure Swift/SwiftUI, no external dependencies • Runs entirely on-device (no server calls for validation) • Deterministic (same input always produces same output) • Auditable (full trace logging for every validation) I'm preparing to submit to the App Store and wanted to ask: Are there specific App Review guidelines I should reference for AI safety claims? Is there interest from Apple in deterministic governance layers for Apple Intelligence integration? Any recommendations for demonstrating safety compliance during review? The app is called Ada, and the engine is open source at: github.com/jaredlewiswechs/ada-newton Happy to share technical documentation or discuss the architecture with anyone interested. See: parcri.net

Hi, I am a new IOS developer, trying to learn to integrate the Apple Foundation Model. my set up is: Mac M1 Pro MacOS 26 Beta Version 26.0 beta 3 Apple Intelligence & Siri --> On here is the code, func generate() { Task { isGenerating = true output = "⏳ Thinking..." do { let session = LanguageModelSession( instructions: """ Extract time from a message. Example Q: Golfing at 6PM A: 6PM """) let response = try await session.respond(to: "Go to gym at 7PM") output = response.content } catch { output = "❌ Error:, \(error)" print(output) } isGenerating = false } and I get these errors guardrailViolation(FoundationModels.LanguageModelSession.GenerationError.Context(debugDescription: "Prompt may contain sensitive or unsafe content", underlyingErrors: [Asset com.apple.gm.safety_embedding_deny.all not found in Model Catalog])) Can you help me get through this?

When using the imagePlaygroundSheet modifier in SwiftUI, the system presets an image playground in a fixed size. Especially on macOS, this modal is rather small and doesn't utilize the available space efficiently. Is there a way to make the modal bigger, or allow the user to resize the dialog? I tried presentationDetents, but this would need to be applied to the content of the sheet, which is provided by the system... I guess this question applies to other system-provided sheets like the photo picker as well.

Hello, I am interested in using jax-metal to train ML models using Apple Silicon. I understand this is experimental. After installing jax-metal according to https://developer.apple.com/metal/jax/, my python code fails with the following error JaxRuntimeError: UNKNOWN: -:0:0: error: unknown attribute code: 22 -:0:0: note: in bytecode version 6 produced by: StableHLO_v1.12.1 My issue is identical to the one reported here https://github.com/jax-ml/jax/issues/26968#issuecomment-2733120325, and is fixed by pinning to jax-metal 0.1.1., jax 0.5.0 and jaxlib 0.5.0. Thank you!

Recently, I'm trying to deploy some third-party LLM to Apple devices. The methodoloy is similar to https://github.com/Anemll/Anemll. The biggest issue I'm having now is the runtime memory usage. When there are multiple functions in a model (mlpackage or mlmodelc), the runtime memory usage for weights is somehow duplicated when I load all of them. Here's the detail: I created my multifunction mlpackage following https://apple.github.io/coremltools/docs-guides/source/multifunction-models.html I loaded each of the functions using the generated swift class: let config = MLModelConfiguration() config.computeUnits = MLComputeUnits.cpuAndNeuralEngine config.functionName = "infer_512"; let ffn1_infer_512 = try! mimo_FFN_PF_lut4_chunk_01of02(configuration: config) config.functionName = "infer_1024"; let ffn1_infer_1024 = try! mimo_FFN_PF_lut4_chunk_01of02(configuration: config) config.functionName = "infer_2048"; let ffn1_infer_2048 = try! mimo_FFN_PF_lut4_chunk_01of02(configuration: config) I observed that RAM usage increases linearly as I load each of the functions. Using instruments, I see that there are multiple HWX files generated and loaded, each of which contains all the weight data. My understanding of what's happening here: The CoreML framework did some MIL->MIL preprocessing before further compilation, which includes separating CPU workload from ANE workload. The ANE part of each function is moved into a separate MIL file then compile separately into a HWX file each. The problem is that the weight data of these HWX files are duplicated. Since that the weight data of LLMs is huge, it will cause out-of-memory issue on mobile devices. The improvement I'm hoping from Apple: I hope we can try to merge the processed MIL files back into one before calling ANECCompile(), so that the weights can be merged. I don't have control over that in user space and I'm not sure if that is feasible. So I'm asking for help here. Thanks.

Posting a follow up question after the WWDC 2025 Machine Learning AI & Frameworks Group Lab on June 12. In regards to the on-device API of any of the AI frameworks (foundation model, vision framework, ect.), is there a response condition or path where the API outsources it's input to ChatGPT if the user has allowed this like Siri does? Ignore this if it's a no: is this handled behind the scenes or by the developer?

I’m developing an activity classifier that I’d like to input using the JSON format of CoreMotion data. I am getting the error: Unable to parse /Users/DewG/Downloads/Testing/Step1/Testing.json. It does not appear to be in JSON record format. A SequenceType of dictionaries is expected I've verified that the format I am using is JSON via various JSON validators, so I am expecting I'm just holding it wrong. Is there an example of a JSON file with CoreMotion data that I can model after?

Hi everyone Im currently developing an object detection model that shall identify up to seven classes in an image. While im usually doing development with basic python and the ultralytics library, i thought i would like to give CreateML a shot. The experience is actually very nice, except for the fact that the model seem not to be using any ANE or GPU (MPS) for accelerated training. On https://developer.apple.com/machine-learning/create-ml/ it states: "On-device training Train models blazingly fast right on your Mac while taking advantage of CPU and GPU." Am I doing something wrong? Im running the training on Apple M1 Pro 16GB MacOS 26.1 (Tahoe) Xcode 26.1 (Build version 17B55) It would be super nice to get some feedback or instructions. Thank you in advance!

When using CoreML for VAE model prediction, the prediction result shows a distorted display with no error messages. How can this issue be addressed?

I am excited to share that I have developed a Metal kernel for Flash Attention that eliminates race conditions and fully leverages Apple Silicon’s shared memory and registers. This kernel can dramatically accelerate training of transformer-based models. Early benchmarks suggest that models which previously required months to train could see reductions to just a few hours on Apple hardware, while maintaining numerical stability and accuracy. I plan to make the code publicly available to enable the broader community to benefit. I would be happy to keep you updated on the latest developments and improvements as I continue testing and optimizing the kernel. I believe this work could provide valuable insights for Apple’s machine learning research and products.

I'm downloading a fine-tuned model from HuggingFace which is then cached on my Mac when the app first starts. However, I wanted to test adding a progress bar to show the download progress. To test this I need to delete the cached model. From what I've seen online this is cached at /Users/userName/.cache/huggingface/hub However, if I delete the files from here, using Terminal, the app still seems to be able to access the model. Is the model cached somewhere else? On my iPhone it seems deleting the app also deletes the cached model (app data) so that is useful.

Using highly optimized Metal Shading Language (MSL) code, I pushed the MacBook Air M2 to its performance limits with the deformable_attention_universal kernel. The results demonstrate both the efficiency of the code and the exceptional power of Apple Silicon. The total computational workload exceeded 8.455 quadrillion FLOPs, equivalent to processing 8,455 trillion operations. On average, the code sustained a throughput of 85.37 TFLOPS, showcasing the chip’s remarkable ability to handle massive workloads. Peak instantaneous performance reached approximately 673.73 TFLOPS, reflecting near-optimal utilization of the GPU cores. Despite this intensity, the cumulative GPU runtime remained under 100 seconds, highlighting the code’s efficiency and time optimization. The fastest iteration achieved a record processing time of only 0.051 ms, demonstrating minimal bottlenecks and excellent responsiveness. Memory management was equally impressive: peak GPU memory usage never exceeded 2 MB, reflecting efficient use of the M2’s Unified Memory. This minimizes data transfer overhead and ensures smooth performance across repeated workloads. Overall, these results confirm that a well-optimized Metal implementation can unlock the full potential of Apple Silicon, delivering exceptional computational density, processing speed, and memory efficiency. The MacBook Air M2, often considered an energy-efficient consumer laptop, is capable of handling highly intensive workloads at performance levels typically expected from much larger GPUs. This test validates both the robustness of the Metal code and the extraordinary capabilities of the M2 chip for high-performance computing tasks.

Here's the result: Very weird.

Hi, guys. I'm writing about Apple Intelligence and I reached the point I have to explain App Intent Domains https://developer.apple.com/documentation/AppIntents/app-intent-domains but I noticed that there is a note explaining that these services are not available with Siri. I tried the example provided by Apple at https://developer.apple.com/documentation/AppIntents/making-your-app-s-functionality-available-to-siri and I can only make the intents work from the Shortcuts App, but not from Siri. Is this correct. App Intent Domains are still not available with Siri? Thanks

Hi! I'm trying to use the ImagePlayground API in SwiftUI with the .imagePlaygroundSheet modifier. However, when the sheet is shown (in the preview or in the simulator) it displays the following message: "Image Playground is not available. Image Playground is not available on this iPhone.". I'm using an iPhone 16 Pro with iOS 18.3.1 in the Xcode (16.2) Simulator. Anyone else having this problem? How can I fix it?

Hi all, I'm tuning my app prediction speed with Core ML model. I watched and tried the methods in video: Improve Core ML integration with async prediction and Optimize your Core ML usage. I also use instruments to look what's the bottleneck that my prediction speed cannot be faster. Below is the instruments result with my app. its prediction duration is 10.29ms And below is performance report shows the average speed of prediction is 5.55ms, that is about half time of my app prediction! Below is part of my instruments records. I think the prediction should be considered quite frequent. Could it be faster? How to be the same prediction speed as performance report? The prediction speed on macbook Pro M2 is nearly the same as macbook Air M1!

Hi, I am modifying the sample camera app that is here: https://developer.apple.com/tutorials/sample-apps/capturingphotos-camerapreview ... In the processPreviewImages, I am using the Vision APIs to generate a segmentation mask for a person/object, then compositing that person onto a different background (with some other filtering). The filtering and compositing is done via CoreImage. At the end, I convert the CIImage to a CGImage then to a SwiftUI Image. When I run it on my iPhone, it works fine, and has not crashed. When I run it on the iPhone with the debugger, it crashes within a few seconds with: EXC_BAD_ACCESS in libRPAC.dylib`std::__1::__hash_table<std::__1::__hash_value_type<long, qos_info_t>, std::__1::__unordered_map_hasher<long, std::__1::__hash_value_type<long, qos_info_t>, std::__1::hash, std::__1::equal_to, true>, std::__1::__unordered_map_equal<long, std::__1::__hash_value_type<long, qos_info_t>, std::__1::equal_to, std::__1::hash, true>, std::__1::allocator<std::__1::__hash_value_type<long, qos_info_t>>>::__emplace_unique_key_args<long, std::__1::piecewise_construct_t const&, std::__1::tuple<long const&>, std::__1::tuple<>>: It had previously been working fine with the debugger, so I'm not sure what has changed. Is there a difference in how the Vision APIs are executed if the debugger is attached vs. not?