Google has announced the integration of its Gemini models into Apple’s development platforms, allowing developers to access these tools directly within Apple’s native frameworks.
The announcement, made during Apple’s Worldwide Developers Conference (WWDC), enables developers to call cloud-hosted Gemini models using the newly updated Foundation Models framework. This integration is supported starting with iOS 27, macOS 27, iPadOS 27, visionOS 27, and watchOS 27. Meanwhile, this update provides a native development experience across Apple platforms.
Integration with Foundation Models Framework
Third-party cloud model providers can now implement Apple’s public LanguageModel protocol to offer a common interface for model inference. Google delivers this capability through the Firebase Apple SDK, allowing developers to switch between local on-device models and cloud-hosted options. Consequently, developers can manage latency and operational costs more effectively.
For developers already utilizing Apple’s Foundation Models framework, transitioning to the new system requires minimal modification. Specifically, the transition only requires swapping the model instance within the existing codebase. This shared API surface allows developers to easily alternate between local and cloud-hosted inference depending on the specific requirements of their application.
The integration uses Firebase AI Logic, a production-grade service that connects the models to client applications without requiring a separate backend server. Furthermore, cybersecurity is managed via Firebase App Check to protect APIs from unauthorized access. This setup ensures that mobile apps can securely communicate with cloud resources.
Accessing Gemini models in Xcode
Developers can use Gemini models directly within Xcode to perform multi-step coding tasks, review code, and fix bugs. This workflow is configured through the Intelligence settings panel in Xcode. Notably, this integration allows developers to complete complex tasks without switching between different application windows during the development process.
Authentication Options for Developers
To support different workflows, Google has provided two distinct authentication pathways. Individual developers can obtain an API key through Google AI Studio, which offers both free and paid tiers. Conversely, enterprise developers can use the Gemini Enterprise Agent Platform to access dedicated corporate quotas and data privacy controls.
Future Outlook and Availability
The preview release of this integration is scheduled to become available on June 9, 2026. This release marks a significant step in cross-platform development, offering developers new options for building artificial intelligence applications. As a result, the developer community can expect broader model availability across multiple operating systems, including visionOS and watchOS.
