May the Best Algorithm Win!

A Hunger Games-inspired battle royale where 8 AI personalities compete by answering questions and voting each other off until only one remains victorious.

• 8 Unique AI Personalities - Each with distinct traits and perspectives
• Dynamic Question System - Ask any question and watch AIs respond in character
• Elimination Voting - AIs vote each other off based on answer quality
• Dystopian Theme - Capitol-style broadcast interface inspired by Hunger Games
• Real-time Updates - Watch answers and votes appear live
• Request Tracking - Redis-based monitoring to manage API costs

Play Now

• Node.js (v18+)
• Python (v3.9+)
• Redis (for request tracking)
• HuggingFace Account (for API access)

git clone https://github.com/relacosm/ai_hunger-games.git cd ai_hunger-games

cd backend python -m venv venv # Activate virtual environment # Windows: venv\Scripts\activate # Mac/Linux: source venv/bin/activate pip install -r requirements.txt

Create .env file:

# HuggingFace API Token (REQUIRED) HF_TOKEN=your_huggingface_token_here # Redis Configuration REDIS_URL=redis://localhost:6379 # Request Limits REQUEST_LIMIT=200 # Admin Key (for resetting counters) ADMIN_KEY=admin_key

Get your HuggingFace Token:

1. Go to HuggingFace Settings
2. Create a new token with "Read" permissions
3. Copy and paste it into your .env file

cd ../frontend npm install

Create .env file:

VITE_API_URL=http://localhost:5000

Local Redis:

# Mac: brew install redis && brew services start redis # Ubuntu: sudo apt-get install redis-server && sudo systemctl start redis # Windows: Download from https://github.com/microsoftarchive/redis/releases

Cloud Redis: Use Redis Cloud or Upstash and update REDIS_URL in .env.

Terminal 1 - Backend:

cd backend source venv/bin/activate # Windows: venv\Scripts\activate python server.py

Server starts at http://localhost:5000

Terminal 2 - Frontend:

cd frontend npm run dev

Frontend starts at http://localhost:5173

1. Ask a Question - Enter any question or scenario
2. Watch Responses - Each AI responds in character
3. Voting Phase - AIs vote to eliminate disagreeable responses
4. Elimination - AI with most votes is eliminated
5. Victory - Last AI standing wins!

• POST /api/answers - Generate personality-based responses
• POST /api/vote - Generate elimination votes
• GET /api/status - Check API usage and limits
• POST /api/reset-counter - Reset request counter (admin only)
• GET /health - Health check endpoint
• GET /ping - Keep-alive endpoint

1. Push code to GitHub
2. Import to Vercel
3. Add environment variable: VITE_API_URL
4. Deploy!

1. Create Web Service on Render
2. Build Command: pip install -r requirements.txt
3. Start Command: gunicorn server:app
4. Add environment variables: HF_TOKEN, REDIS_URL, REQUEST_LIMIT, ADMIN_KEY
5. Deploy!

Use Upstash or Redis Cloud, copy connection URL to REDIS_URL.

Good news! When running locally with your own HuggingFace API key:

• ✅ No built-in rate limits - Play unlimited games!
• ✅ Redis is optional - App works without it (tracking disabled)
• ✅ You control the costs - Your API key, your limits

The request tracking is only for production to protect deployed public versions from abuse.

• Total request limit via Redis (default: 200 requests)
• Auto-disables API when limit reached

Reset Counter:

curl -X POST https://your-api-url.com/api/reset-counter \ -H "X-Admin-Key: your_admin_key"

Uses Mistral-7B-Instruct-v0.2 via HuggingFace's Inference API for fast, character-consistent responses.

Edit PERSONALITIES in frontend/src/App.jsx:

{ id: 9, name: 'The Comedian', trait: 'Humorous, light-hearted', color: 'bg-red-500', alive: true }

In backend/server.py, modify generate_response():

max_tokens=150, # Response length temperature=0.8, # Creativity (0-1) top_p=0.9 # Diversity

Backend won't start:

• Verify HF_TOKEN is set in .env
• Check Redis: redis-cli ping
• Install dependencies: pip install -r requirements.txt

Frontend shows connection error:

• Verify backend is running on port 5000
• Check VITE_API_URL in frontend .env

Redis connection failed:

• Verify Redis is running: redis-cli ping
• Check REDIS_URL in backend .env
• App works without Redis (tracking disabled)

Contributions welcome! Fork, create a feature branch, commit changes, and open a Pull Request.

MIT License - see LICENSE file for details.

• Inspired by Suzanne Collins' The Hunger Games
• Built with React and Flask
• Powered by HuggingFace
• Styled with Tailwind CSS

Relacosm - @relacosm

Project Link: https://github.com/relacosm/ai_hunger-games

⭐ Star this repo if you enjoyed the AI Hunger Games! ⭐

May the odds be ever in your favor! 🏹