Google brings Gemini to Apple developers via Foundation Models framework and Xcode at WWDC 2026

Google announced at Apple's WWDC 2026 that Gemini cloud models are now directly accessible to Apple developers through Apple's Foundation Models framework and built into Xcode. Developers building for iOS 27, macOS 27, iPadOS 27, and visionOS can call Gemini via the same API used for on-device Apple models, using a single parameter change to swap between local and cloud inference.

What's new

Google's announcement confirmed that "Apple developers can now securely call cloud-hosted Gemini models using the Foundation Models framework, and access Gemini in Xcode."

Key details of the integration:

• Firebase AI Logic SDK: Lets developers call cloud-hosted Gemini from iOS, macOS, iPadOS, and visionOS apps without building or maintaining a separate backend server
• Unified API surface: The SDK aligns Gemini's call signature with Apple's local Foundation Models API, enabling developers to "easily swap between local and cloud inference to fit your use case"
• Xcode integration: Gemini is embedded directly in Xcode to help developers "review code, fix bugs and build new features faster"
• Authentication tiers: Individual developers use self-serve API keys; enterprise customers access the Gemini Enterprise Agent Platform for dedicated quotas and privacy controls
• Availability: The Foundation Models framework integration launches as a preview release alongside iOS 27 and macOS 27

Context

Apple's Foundation Models framework is the new developer-facing abstraction for on-device AI, introduced at WWDC 2026. It standardizes how apps call small, locally-running models — keeping latency low and costs zero for common tasks. By opening the framework to Google's Gemini at the cloud level, Apple is extending Foundation Models from a purely on-device interface into a hybrid inference platform.

The Apple-Google collaboration on AI runs deeper than this announcement. Separately at WWDC 2026, NVIDIA Blackwell GPUs are running confidential inference in Apple's Private Cloud Compute infrastructure for Apple Foundation Models built in collaboration with Google — underscoring that the two companies are working together across multiple AI infrastructure layers.

Why it matters

For Google, this is a significant distribution win. Gemini embedded in Xcode places it in front of the entire Apple developer ecosystem as the first — and currently only — cloud AI option available through the Foundation Models framework. The Firebase AI Logic SDK removes the friction that historically slowed cloud AI adoption among mobile developers: no backend infrastructure to build, no separate API integration to maintain.

For the broader developer ecosystem, the integration signals that Apple's Foundation Models framework is designed to be extensible at the cloud tier. That architecture could eventually open the door to other AI providers offering compatible endpoints — though for WWDC 2026, Gemini is the only cloud option on the table.