WWDC 2026 has come and gone with a average quantity of fanfare, at the very least in comparison with earlier years. When you simply watched the Keynote on Monday, you noticed 3 predominant themes, however for one among them, there was lots sitting beneath the floor.
Platform Enhancements
Alongside the discharge of macOS Golden Gate (macOS 27), the remainder of the “27” releases have been introduced, conserving with the 12 months naming conference began in final 12 months’s WWDC. While you assume “platform enhancements”, these which have been with Apple for some time could also be having flashbacks to Snow Leopard, which was a model of OS X (which is what macOS glided by again then), that was largely centered on fixing bugs and enhancing efficiency, as a substitute of recent options.
This 12 months, Apple is giving the entire “27” OSes a little bit of polish up entrance – specifically with a consumer adjustable glass transparency in Liquid Glass, together with different tweaks – and a LOT of efficiency tune ups beneath the hood. We’ll need to see how these all translate to actual life, particularly on older gadgets (with help going again to iPhone 11!), however they quote some fairly good enhancements. Simply being able to maintain utilizing iMessage whereas a big footage transmits is value it for me!
Belief and Security
For anybody who manages their youngster’s Apple gadgets, the in-built instruments like Display Time have been…. okay? Like many issues all through the OS, these controls wanted their very own polish, bug fixes, and enhancements.
Mother and father can now extra finely tune the apps, web sites, contacts and instances that kids can use their gadgets. The “Ask-to-buy” idea has been carried over to asking for entry to web sites and contacts, which is a pleasant function. The additional constancy additionally means that you can present your youngster entry to extra issues as they become old.
Apple Intelligence and Siri AI
The remainder of the keynote, the vast majority of the Platforms State of the Union and a really, very giant variety of the developer movies launched this week revolved across the enhancements round Siri AI, the ever increasing AI frameworks (a few of which are actually open sourced) and what underpins all of it, Apple Intelligence and App Intents.
They keynote exhibits what Siri AI can do for the top consumer, however how do you allow these options in your app? Let’s check out a few of the highlights within the “AI & Machine Studying” part of this 12 months’s WWDC:
Xcode Enhancements for AI
The introduction of AI capabilities beginning in Xcode 18 and persevering with into Xcode 26 confirmed some first rate efforts to construct help into the IDE. Even this 12 months, the introduction of agentic capabilities allowed builders to take a extra palms off strategy to coding.
Xcode 27, although? Wow.
Not solely is it simpler than ever to allow fashions, together with customized fashions operating in your native machine, you possibly can supercharge their talents with in-built brokers particularly designed for Apple workflows. All of it begins with Assistant Conversations, which are actually introduced in their very own window, to allow them to be organized similar to some other editor.
Let’s say you need to Localize your app to French. Earlier than, this required discovering translations for all your consumer going through strings, constructing a strings file, ensuring all of these translations made sense not solely to the native language however to your app. It was lots. Now you simply want one thing like this.
Then you possibly can sit again and watch Xcode localize the app for you! Even higher, you possibly can multitask. Whereas that localization is happening, you would begin one other dialog to localize into one other language, and even add Accessibility labels. Now there’s no excuse to allow your app to achieve as many individuals as potential world wide.
App Schemas are usable!
App Schemas have been round in some form for a number of years now. They have been initially supposed to facilitate what we’re simply now seeing on this 12 months’s WWDC (keep in mind when Apple Intelligence first got here out… and the way it actually didn’t?). App Schemas can help you adorn your App Entities and App Intents so Apple Intelligence know the way to finest use them. Apple’s studying fashions have been skilled on key options of iOS resembling calendars, mail messages, and photographs, so in case your App Entities or App Intents use these options, you’ll get tight integration with Siri AI, permitting you do enjoyable issues like the varied multi-step conversations seen within the Keynote.
The important thing factor to recollect right here is that in case your Entities and Intents *don’t* conform to one among these schemas, you received’t get the total advantages of Siri AI, however you possibly can nonetheless use them in different areas like Shortcuts or the Motion button.
App Intents API enhancements
Since App Schemas are literally helpful now, utilizing them when making App Intents is less complicated as effectively! At its core, App Intents haven’t change lots since final 12 months. There are some good APIs which have been included to help offering values to intents, however in case you have your EntityQuerys setup, you’re in all probability good to go. Xcode’s autocomplete when working with App Schemas makes filling out the App Entities and App Intents a breeze. 
Core AI
For these of you into crafting or superb tuning the fashions you utilize in your app, a brand new framework referred to as Core AI is the place to go. This new inference framework covers the total mannequin lifecycle, profiting from the {hardware} in your machine, and comes with Swift and Python APIs, alongside a developer toolchain inside Xcode and Devices.
Basis fashions enhancements
The Basis fashions framework that Apple launched final 12 months has seen some main enhancements. Along with offering a extra sturdy on machine mannequin and the flexibility to make use of Google-powered Personal Cloud Compute (which has a robust reasoning mannequin), now you can deliver any mannequin into your LanguageModelSession.
// On-device Apple Basis Mannequin
let mannequin = SystemLanguageModel()
// let mannequin = PrivateCloudComputeLanguageModel()
// let mannequin = MLXLanguageModel(modelID: "mlx-community/my-model")
let session = LanguageModelSession(mannequin: mannequin)
You’ll be able to even use MLX, MLX-LM, and the brand new MLX-LM Server to run native fashions in your native machine, similar to instruments resembling LM Studio.
Dynamic Profiles can help you program a number of states on your app, even permitting you to jot down a device to modify between them primarily based on what the consumer is doing in your app. Every profile appears like this:
Profile {
Directions {...}
TaskSpecificTool()
}
Every of those instruments will be setup with completely different fashions, completely different ranges of reasoning, and completely different device help, giving your consumer simply what they want when utilizing your app, all with out them lifting a finger.
