Gemini-Guess: Our Hackathon MVP for the Gemini 3 Hackathon
We’re excited to announce that Q.Z.L Corp is joining the Gemini 3 Hackathon with our project Gemini-Guess!
You can also see our preview on Devpost here: https://devpost.com/software/gemini-guess.
What is Gemini-Guess?
Gemini-Guess is an open-ended game where players challenge Gemini 3 to guess any idea, object, or concept. Players can give clues via text, voice, or video, and Gemini 3 responds with deep, multi-turn reasoning, showing its thought process step by step.
It’s designed to test curiosity, creativity, and AI reasoning in a fun and interactive way.
Inspiration
The idea came from how much people, especially kids, love guessing games. We wanted to explore how far AI can reason in open-ended scenarios. Gemini 3’s deep reasoning capabilities made it the perfect choice for this experiment.
What it does
- Players think of any concept, object, or idea — or say/show it visually.
- Gemini 3 tries to guess it through multi-turn reasoning.
- Tracks rounds, scores, and allows replay of the AI’s thought process.
- Supports text, voice, and video input for natural interactions.
Try it here: Gemini-Guess MVP
How we built it
- Next.js for the frontend — fast for hackathon prototyping.
- Gemini 3 API powers reasoning and multi-turn guesses.
- Web APIs handle microphone and camera input for voice/video clues.
- Game state stored in memory, ready for leaderboards in the future.
- Replays visualized and available with audio for judges and players.
Challenges we ran into
- Designing prompts that handle unlimited concepts.
- Handling voice and video input with low latency.
- Keeping the game fun but fair, letting AI sometimes fail in interesting ways.
- Managing multi-turn reasoning in a concise replayable format.
Accomplishments
- Showcased Gemini 3’s reasoning depth and nuance.
- Created a replayable AI thought process for users and judges.
- Supported multimodal input (text, voice, video) in a single app.
- Made an interactive experience that’s accessible for everyone.
What we learned
- Gemini 3 can handle deep, multi-turn reasoning across open-ended challenges.
- Multimodal input makes AI interaction more natural.
- Prompt design is both an art and a science.
- Watching AI think in real time is as fun as playing the game.
Next Steps for Gemini-Guess
- Add leaderboards and global scoring.
- Expand video/voice AI interaction, possibly live-stream challenges.
- Optimize prompts for faster AI guesses without losing nuance.
- Build a mobile-friendly React Native version for on-the-go play.
We can’t wait for everyone to try it and see how far Gemini 3 can go in guessing your ideas. Check it out here: https://gemini-guess.qz-l.com/
Stay tuned for updates as we improve the game during the hackathon!