What’s new in AI & machine learning

The Foundation Models framework is a native Swift API that gives you direct access to the same on-device model that powers Apple Intelligence. You can now work with any language model, including Apple Foundation Models, cloud models like Claude and Gemini, or any other provider that conforms to the Language Model protocol.

Multimodal prompts let you pass images alongside text so your app can reason about visual content, and Vision framework tools like OCR and barcode readers are available for your model to call directly, all on-device. Dynamic Profiles let you swap models, tools, and instructions on the fly, so your app’s behavior can adapt within a continuous session.

If your app has fewer than 2 million total first-time App Store downloads, you can access the latest Apple Foundation Model on Private Cloud Compute, with unparalleled privacy protections. And with the Evaluations framework, you can verify that your AI features behave correctly across dynamic conditions, going beyond what unit tests alone can catch.

Core AI is a new framework built directly into the OS and purpose-built for Apple Silicon, providing the best way to bring your own models on-device — complete with supporting tools and technologies. A modern, memory-safe Swift API lets you load, specialize, and run AI models entirely on-device, keeping user data private and your apps responsive, with zero server dependencies and zero token costs. Models are automatically specialized for the hardware they run on, with ahead-of-time compilation support for quick load times. Fine-grained control over inference memory, zero-copy data paths, and stateful execution give you the performance you need to run everything from compact vision models to large-scale generative AI across all Apple platforms.

MLX, the open source array framework for Apple silicon, is faster than ever and more capable of experimenting with, training, and fine-tuning large language models. This year it gains support for Metal 4 and GPU Neural Accelerators for even greater performance, and you can now scale training across multiple Macs using RDMA over Thunderbolt — making serious model research and fine-tuning accessible right from your desk.

The new Evaluations framework gives you the tools to test your prompts and validate that your intelligence-powered features work reliably before they reach users. Build confidence in your AI-driven experiences with systematic evaluation built directly into your development workflow, so you can ship knowing your features perform as intended.