Join us for a one-day, AI-first hackathon where developers, students, and innovators build impactful applications using Gemma open models.
Welcome to the Gemma 4 Hackathon Sprint, a fast-paced, 1-day in-person hackathon where you'll get hands-on with the most capable open models from Google DeepMind. This hackathon is not about writing polished, production-ready code. It is a sprint to experiment, learn, break things, and build great solutions to real-world problems in just a few hours. You have access to the newly released Gemma 4 models. Team up, brainstorm a wild idea, hack a prototype, and pitch before the clock runs out. Bring your laptop, your boldest ideas, and let's build something amazing together!
With the release of Gemma 4, a new family of open models is officially in your hands. Leverage local frontier intelligence, native function calling, and multimodal understanding to tackle core developer challenges.
Tracks * The GenAI for Good Track: Best application addressing high-impact challenges in healthcare, agriculture, or civic engagement using Gemma. * The Autonomous Agent Track: Best use of Gemma's native function calling to build local AI agents that interact with external APIs or tools.
Your Goal Build an AI application powered by Gemma 4 that solves a real-world problem. Participants are encouraged to leverage Gemma 4's language understanding, reasoning, multimodal, and agentic capabilities to create innovative solutions across any domain. Projects should demonstrate meaningful use of Gemma as the primary intelligence behind the solution and deliver a functional prototype by the end of the hackathon.
Submission Requirements A valid submission must contain the following: Kaggle Writeup, Attached Public Code Repository, Attached Live Demo (or Clonable Notebook).
Awards * The GenAI for Good Track: $1,600 total ($1,000 First Prize, $600 Second Prize). * The Autonomous Agent Track: $400 total (Track Winner $400).
Evaluation Applications will be evaluated based on: Gemma Integration (30%), Innovation & Impact (30%), Functionality (20%), Presentation & Writeup (20%). Projects that do not make meaningful use of Gemma as a core component of the solution may be disqualified from judging.