Apple refines Genmoji creation with iterative editing in iOS 27
MacRumors discovers updated Genmoji workflow that allows users to tweak attributes without restarting the generation process.
Apple is introducing a significant update to its Genmoji creation tools in iOS 27, allowing users to refine generated emojis through an iterative editing process. According to a discovery by MacRumors, the new experience includes a "Describe a change" text interface that enables users to modify attributes such as colour and added elements without having to restart the generation process from scratch.
The updated workflow provides multiple entry points for creation, allowing users to generate a Genmoji from an existing emoji, an image from the Photos app, a person tagged in their gallery, or a written text prompt. Once an initial result is generated, users can utilise the new text interface to request specific alterations. For instance, a user could generate an orange cat with an umbrella and then prompt the system to change the cat to a calico, avoiding the need to regenerate the entire image.
This contrasts with the current Genmoji creator, which requires users to start a brand new process if they are unsatisfied with the initial output. The new system reportedly produces more consistent results, with a default style of cute, 3D cartoon emojis. Users can also adjust the aesthetic through new style options, including drawings and sketches, offering greater flexibility in the final design.
Apple did not explicitly detail this specific update during its WWDC announcements for iOS 27 and Apple Intelligence. However, the company included a statement in its Apple Intelligence reveal noting that "Genmoji quality is even better and allows users to describe the changes they want to make." The confirmation of these UI changes comes from MacRumors' examination of the iOS 27 developer build.
Apple Intelligence features, including the enhanced Genmoji tools, are scheduled to roll out to users this fall. The discovery highlights Apple's ongoing effort to refine its generative AI capabilities, moving beyond basic generation to offer more precise control over the output for end users.
