Reportedly, code digging in the first developer betas of macOS Tahoe 26.1, iPadOS 26.1, and iOS 26.1 revealed that Apple will introduce new framework support for Image Playground, adding features like "estimated latency" and provider identification, paving the way for integrating more third-party models.
According to the official introduction, Image Playground is Apple's AI-powered image generation app that generates illustrations based on text prompts. In the first developer betas of macOS Tahoe 26.1, iPadOS 26.1, and iOS 26.1, Apple updated Image Playground's underlying framework.
Originally, the app generated illustrations solely based on Apple's own models. Later this year, it introduced OpenAI's ChatGPT, providing more image styles and direct model calls. In the latest beta code, Apple has added an "estimated latency" metric and provider identification to Image Playground to evaluate the performance of different models.
Apple hasn't yet announced the specific models it will add, but the media outlet suggests Google's Gemini 2.5 Flash Image (also known as Nano Banana) is a leading candidate. This model has gained popularity for its ability to realistically reproduce human features, propelling the Google Gemini app to the top of the App Store in multiple countries. In future versions, users may experience even higher quality and richer styles of AI-generated images.
