Apple Photorealistic AI Image Generation Launches in Image Playground with New Any Style Engine and Spatial Reframing Features for iOS 27
When Apple first brought generative creative tools to the world, output was limited to stylised sketches, animations, and illustrative avatars. In a massive volte face, this was overturned at WWDC 2026 with Apple launching photorealistic AI image generation directly into the refreshed Image Playground app.
As part of the general update with iOS 27 and macOS 27, the refreshed creative tools received an Any Style option that works directly inside the app. It is powered by next generation diffusion models running on Apple's private cloud servers which can be commanded using descriptive natural language prompts and allow the generation of hyper realistic photos, mockups and even presentation layouts.
The fact that these server based models are highly resource hungry means Apple will cap the daily usage of photorealistic generation at 100 requests per day although those using any iCloud+ tier of subscription will see increased daily access levels, as well as greater access to security video processing.
The Photos app is also getting a significant generative tool known as Spatial Reframing, which is powered by context aware machine learning, and will allow you to procedurally extend the outer boundaries of your photographs or modify their interior structure long after the physical shutter button has been pressed making poorly focused or tight shots more amenable to post editing and creating a panoramic perspective on a landscape.
