Apple WWDC26: Foundation Models gain Gemini support, Xcode 27 gets agentic coding tools

Apple has announced a broad range of developer-focused updates at WWDC26, introducing new AI frameworks, expanded Apple Intelligence capabilities, agentic coding tools in Xcode 27, and platform enhancements designed to make apps faster, more adaptive, and easier to build.

Apple Intelligence and Foundation Models gain major upgrades

Apple is expanding its intelligence frameworks with new capabilities that allow apps to work more closely with Apple Intelligence and Siri AI. Updates to the App Intents framework enable developers to connect their apps to Siri AI features such as personal context understanding, app actions, and onscreen awareness.

At the center of the update is the Foundation Models framework, which now serves as a single native Swift API for accessing AI models.

Key additions include:

  • More powerful on-device AI models
  • Image input support
  • Server-hosted AI models
  • Custom AI skills
  • Support for third-party cloud model providers
  • Dynamic Profiles for updating model behavior in real time

Apple is also opening the framework to third-party cloud model providers through a new LanguageModel protocol. Google Gemini is among the first models available through the Firebase Apple SDK, allowing developers to switch between Apple’s on-device models and cloud-hosted Gemini models using the same API surface. The integration leverages Firebase AI Logic and Firebase App Check, enabling developers to integrate Gemini into their apps without building a separate backend server.

Apple says the next generation of Apple Foundation Models was developed in collaboration with Google and its Gemini models. Developers can also use models such as Gemini, Claude, and other services that implement the LanguageModel protocol.

Developers enrolled in the App Store Small Business Program with fewer than 2 million total first-time App Store downloads can access the next generation of Apple Foundation Models running on Private Cloud Compute at no cloud API cost.

Core AI enables local deployment of custom models

Apple has introduced Core AI, a new framework designed for developers who want to run their own AI models directly on Apple devices.

Optimized for Apple silicon, Core AI takes advantage of unified memory and the Neural Engine to enable deployment of full-scale large language models locally without relying on cloud infrastructure.

Xcode 27 expands agentic coding capabilities

Apple is significantly expanding AI-assisted development in Xcode 27. The new release integrates models and coding agents from Anthropic, Google, and OpenAI directly into the development workflow.

Developers can interact with coding agents through conversations that support:

  • Interactive planning
  • Multi-turn Q&A
  • Markdown rendering
  • Code previews and change visualization

Coding agents can also validate their own work by writing and running tests, experimenting in Playgrounds, checking visual changes through previews, and interacting with apps through the new Device Hub.

Google has also integrated Gemini directly into Xcode. Developers can enable it through the Intelligence settings panel to review code, fix bugs, and build new features without leaving the development environment.

Authentication options include:

  • Google AI Studio for individual developers, offering both free and paid tiers
  • Gemini Enterprise Agent Platform for organizations that require dedicated quotas and enterprise privacy controls

Apple is also introducing support for custom plug-ins, Model Context Protocol (MCP), and Agent Client Protocol (ACP), allowing developers to connect external tools and AI agents to Xcode. GitHub and Figma are the first platforms to provide direct integrations.

Beyond agentic coding, Xcode 27 is now Apple silicon-only, occupies 30% less storage space, delivers faster performance, simplifies setup, and introduces a customizable toolbar and a new editor-wide theme system.

Xcode Cloud is now up to twice as fast and adds support for Metal-based applications and visionOS builds, all powered by Apple silicon infrastructure.

SwiftUI, Swift, and design updates

Apple’s Liquid Glass design language continues to evolve with improvements focused on readability, customization, and consistency across platforms. A new transparency slider allows users to adjust the visual appearance while maintaining compatibility with accessibility settings.

SwiftUI now improves responsiveness through faster layout rendering and more efficient state initialization without requiring code changes.

New SwiftUI additions include:

  • Faster layout rendering
  • More efficient state initialization
  • Reorderable containers
  • Spatial Preview framework support

The Spatial Preview framework allows developers to view 3D models from Mac applications spatially when streamed to Apple Vision Pro.

Apple says Swift continues to serve as a unified language for full-stack development, providing the speed, expressiveness, and memory safety needed for mobile applications, embedded firmware, and platforms beyond the Apple ecosystem.

Swift 6.4 introduces:

  • Targeted warning suppression
  • Simplified availability attributes such as “anyAppleOS”
  • Improved compiler diagnostics
New tools for game developers

Apple also announced several updates aimed at game and spatial computing developers.

New game development tools include:

  • Managed Background Assets
  • Steam Asset Converter
  • Game Porting Toolkit 4
  • Official Unity plug-ins
  • Reality Composer Pro 3

Managed Background Assets reduce installation sizes through intelligent localized delivery by automatically downloading language-specific asset packs based on a player’s preferences.

Steam Asset Converter simplifies adapting PC games for Apple platforms, while Game Porting Toolkit 4 introduces open-source AI agent skills that incorporate Apple-specific Metal development best practices.

Apple is also releasing official Unity plug-ins for StoreKit and Background Assets, enabling native In-App Purchases and asset delivery services within existing Unity workflows.

Reality Composer Pro 3 adds Live Previews and Mac Virtual Display support, allowing developers to build and preview complete spatial experiences within a single tool.

Availability

Developer betas for iOS 27, iPadOS 27, macOS 27, watchOS 27, tvOS 27, visionOS 27, and Xcode 27 are available starting today through Apple’s developer program. Apple Intelligence features will be available only in supported regions.

Google said Gemini support for Apple’s Foundation Models framework will be available as a preview release starting today.


Related Post