WWDC26: What’s new in image understanding | Apple
Apple's WWDC26 adds image inputs to Foundation Models and a new Tap to Segment Vision API.
“this year Foundation Models is supporting image inputs”
At WWDC26, Apple's Vision framework team introduced new image-understanding capabilities, including a Tap to Segment API for isolating any object in an image and image input support in the on-device Foundation Models framework. These updates let developers run multimodal LLM analysis and advanced segmentation across iOS and now watchOS, signaling Apple's continued push of on-device generative AI for developers.