Apple to Open On-Device AI Models to Developers at WWDC 2025

Looking to integrate Apple’s powerful AI into your app? At WWDC 2025, Apple is expected to unveil a major update that will open its local AI models to third-party developers. This move is set to empower app creators with access to Apple Intelligence's on-device capabilities, offering features like AI-powered writing tools, image generation, and smart summarization—all without relying on the cloud. By allowing developers to tap into these localized AI models, Apple could reshape how AI is integrated across the iOS ecosystem, delivering faster performance and enhanced privacy for end users.

According to Bloomberg, Apple will provide developers with a new software development kit (SDK) that unlocks smaller, on-device large language models (LLMs)—similar to Google's Gemini Nano AI APIs introduced at Google I/O 2025. This SDK will enable developers to embed Apple’s native AI into their own apps, dramatically increasing functionality while keeping user data secure and processing efficient. While cloud-based model access is off the table for now, this SDK represents a significant step toward Apple’s broader vision of privacy-first, device-native AI.

Currently, developers have limited access to Apple Intelligence features, such as integrating AI Writing Tools or the Image Playground generator. These tools offer capabilities like rewriting text, summarizing content, or generating visual assets using generative AI. However, with the upcoming SDK, developers will gain deeper integration options that make their apps more intelligent and responsive—without needing to build or license their own models like ChatGPT, Grok, or Anthropic.

The announcement is expected to be one of the key highlights of WWDC 2025, which begins on June 9th. Renowned Apple insider Mark Gurman reports that Apple plans to spotlight this shift toward developer AI access alongside the rumored “Solarium” interface redesign. This overhaul will unify iOS, macOS, and iPadOS under a more cohesive design language, one that echoes the look and feel of the Vision Pro operating system.

By opening its local AI models to developers, Apple is not just catching up to Google—it’s doubling down on its core strength: privacy-first innovation. On-device processing ensures sensitive user data never leaves the device, aligning with Apple's long-standing security promises. For developers, it’s a win-win: they get the power of Apple’s AI without compromising on user trust or performance.

This update also signals a broader shift in the AI landscape. With high-performance, energy-efficient chips like Apple Silicon now widespread across devices, more AI workloads can happen directly on the device, reducing reliance on cloud servers and latency. This not only cuts operational costs for developers but also enables real-time, offline capabilities—highly appealing for apps in productivity, healthcare, finance, and education.

As artificial intelligence continues to shape the future of mobile and desktop applications, Apple’s decision to open its on-device AI models could set new standards for how developers harness machine learning in consumer-facing apps. Whether you're building an AI writing assistant, productivity suite, photo editor, or real-time language translator, the SDK will provide a foundation of powerful tools built on Apple's robust AI architecture.

Keep an eye on WWDC 2025 for full details on how to gain access to this SDK and start integrating Apple’s on-device AI features into your next-gen app.