Google is building a way to publish and share AI-generated image and video styles directly from Gemini — according to an APK teardown spotted by Android Authority on October 13, 2026. The feature is not live yet. It’s buried in unreleased code for the Gemini app, flagged as a future capability under active development. No launch date, no rollout schedule, no public beta sign-up. Just strings, UI placeholders, and internal labels pointing to a ‘Publish’ button inside the app’s upcoming redesigned ‘Images’ and ‘Videos’ sections. That’s all we know. And it’s enough to signal a shift: Gemini may soon treat style outputs, not just prompts or final images. As reusable, shareable assets. Think of it like publishing a preset in Photoshop, but for AI aesthetics: lighting, grain, composition bias, even motion cadence in video clips. The teardown doesn’t say whether these published styles will be public by default, require approval, or support attribution. It doesn’t name file formats, licensing terms, or moderation safeguards. It only confirms Google is wiring up the plumbing. The feature appears alongside broader UI reworks, new dedicated tabs for Images and Videos, likely to replace the current unified feed. That suggests Google is treating generative media as first-class output, not just assistant-side garnish. This isn’t about replacing Google Assistant. It’s about making Gemini a platform for repeatable, distributable AI expression. And it arrives while competitors like Meta’s Imagine Studio and Microsoft’s Copilot Image Creator remain strictly one-off tools. No sharing, no reuse, no gallery. Gemini’s move, if shipped, would be the first major AI chat app to let users export and redistribute stylistic DNA. But again: no price, no date, no device requirements. Just code. And a headline that now has to include ‘Gemini AI’, because that’s what people are searching for.