Resources:

The free tier is only available in AI Studio. To get started, access the Build tab of AI Studio to easily and quickly build advanced applications. You can also find existing templates to get started in our Gallery.

📄 Docs

 

📺 Videos

 

Inspiration: 

Need a spark? Start here. Gemini 3 isn’t just a chatbot—it’s a multimodal reasoning engine. It can see, hear, and understand the world in real-time. Don't just build a text interface; build something that senses and reacts.

The launch of Gemini 3 signals the start of the Action Era, transitioning from static chat to autonomous agents that plan and execute complex, long-running tasks. This global hackathon, hosted by Google DeepMind and Devpost challenges you to build high-fidelity AI-powered applications using the Gemini API.

 

We strongly discourage the following types of projects: 

In the Action Era, if a single prompt can solve it, it is not an application. We are looking for orchestrators building robust systems using the Gemini API in Google AI Studio.

  • Baseline RAG: Gemini 3 Pro features a verified 1M token context window. It natively reasons over entire codebases and document sets; simple data retrieval is now a baseline feature.
  • Prompt Only Wrappers: Applications that are merely system prompts in a basic UI.
  • Simple Vision Analyzers: Basic object identification is obsolete. We want to see spatial-temporal video understanding that recognizes cause and effect.
  • Generic Chatbots: Standard bots for nutrition, job screening, or personality analysis.
  • Medical Advice: No projects generating medical or mental health diagnostic advice.

 

Strategic Tracks:

These tracks focus on market potential and technical depth. We encourage you to innovate beyond these prompts to showcase the full potential of Gemini 3 Pro.

🧠 The Marathon Agent: Build autonomous systems for tasks spanning hours or days. Use Thought Signatures and Thinking Levels to maintain continuity and self-correct across multi-step tool calls without human supervision.

  • ☯️ Vibe Engineering: Leverage the Google AI Studio Build tab and Google Antigravity. Build agents that do not just write code but verify it through autonomous testing loops and browser-based verification artifacts.

  • 👨‍🏫 The Real-Time Teacher: Use the Gemini Live API to synthesize live video and audio for adaptive learning.

  • 🎨 Creative Autopilot: Combine Gemini 3 reasoning with Nano Banana Pro for high-precision multimodal creation. Use higher resolution and localized Paint-to-Edit controls to generate professional, brand-consistent assets with legible text

 

FAQs

  1. How does Access to the free tier of Gemini 3 work?