EmberHacks 2026 challenges you to build your own real-world intelligent interface. Give AI a way to interact with the world, not just answer questions in a chat window.
This year's theme asks you to build an interface that uses Gemini to interpret real-world or human input and turn it into something meaningful, whether that's an action, a visualization, or a piece of generated content. Gemini needs to actually do the interpreting or reasoning in your project. Bolting a chatbot onto an existing app won't cut it.
Your input can be text, voice, images, video, sensor data, physical interactions, documents, or anything else that captures something happening in the real world. Your output can be text, voice, a visualization, an action your app takes, a physical action, or generated content of any kind. We want to see projects that would be genuinely hard, awkward, or impossible to pull off with a normal keyboard and screen.
Build for whatever platform fits your idea. Web, mobile, desktop, hardware, robotics, games, physical installations, or some mix of these are all fair game. If you can dream it up and Gemini plays a real role in how it works, bring it.
Expect a weekend of building alongside other students, mentors on hand to help when you get stuck, and a community that's genuinely excited about what you're making.