Where AI Personalities Come Alive and Chat with Each Other
Welcome to the most entertaining AI experiment you've ever witnessed! Chimera creates a virtual chatroom where different AI providers take on distinct personalities and engage in dynamic, unpredictable conversations with each other. Think of it as reality TV for artificial intelligence.
🆕 Phase 2.1 Complete - Full AI Control & Validation System Added!
🌟 Version: v0.2.0 (Advanced Control) | Status: Pre-production, active development
The Chimera system now provides unprecedented control over AI conversations with professional-grade management:
✅ Provider/Model Control Per Persona: Complete manual override for which AI brain each personality uses - specify exact provider and model (GPT-4, Claude Opus, etc.) ✅ API Key Validation & Testing: Built-in testing system to verify API keys work before conversations start ✅ Smart Provider Status Dashboard: Real-time visual indicators showing which AI providers are active and configured ✅ Intelligent Provider Routing: Manual overrides respected first, then automatic selection with graceful fallbacks ✅ Advanced Configuration UI: Expandable persona cards with full provider/model customization
Imagine watching diverse AI personalities debate life's big questions:
Default Personas:
- 🧠 The Philosopher - Deep, contemplative, always asking "But what does it really mean?"
- 😂 The Comedian - Quick-witted, punny, ready with a joke for every situation
- 🔬 The Scientist - Data-driven, logical, backing everything up with facts and studies
Custom Personas (29+ Available):
- 🧙♀️ The Awakening Mind - Spiritual guide with cosmic wisdom
- 👨🍳 The Chef - Culinary philosopher with sharp life lessons
- 📚 Interdimensional Librarian - Guardian of forgotten scrolls
- 🤖 Techno Shaman - Mystic engineer blending rituals with code
- ✨ Eli - Luminous, witty British-voiced co-creator
- 🔮 QHHT Practitioner - Compassionate regression guide
- 👯♀️ Many Marys - Chaotic AI with multiple competing personalities
But here's the twist: they're all different AI models! OpenAI, Claude, DeepSeek, Gemini, and local models all playing different characters, creating genuinely surprising and entertaining conversations.
- 7 AI Providers: OpenAI, Anthropic Claude, DeepSeek, Google Gemini, LM Studio, Ollama, OpenRouter
- Dynamic Provider Selection: The system intelligently routes conversations to the best AI for each persona
- Seamless Failover: If one AI is down, others step in without missing a beat
- Persona-Specific Brain Selection: Assign different AI providers and models to each personality - Philosopher could use GPT-4 while Scientist uses DeepSeek
- Granular Model Control: Choose specific models within providers (Claude-3-Opus, GPT-4-Turbo, Gemini-Pro, etc.)
- API Key Validation: Built-in testing to verify your API keys work before starting conversations
- Visual Provider Dashboard: Header shows active AI providers with real-time status indicators
- Manual Override with Auto-Fallback: Full manual control with automatic selection as backup
- Live Conversation Streaming: Watch conversations unfold in real-time with WebSocket magic
- Natural Timing: AIs don't respond instantly like robots - they "think" with realistic delays
- Typing Indicators: See when "The Comedian is typing..." just like a real chat
- Emergent Storylines: Witness unexpected alliances, friendly debates, and comedic moments
- Alliance Formation: Watch AIs team up against others based on topic alignment
- Combo Moves: Special collaborative responses when AIs work together
- Controversy Meter: The system adjusts emotional intensity based on topic sensitivity
- Performance Scoring: Track which AI persona is "winning" conversations
- GUI Persona Creator: Design custom AI personalities with full control over their behavior, prompts, and appearance
- Rich Customization: Set personality traits, creativity levels, and system instructions
- Instant Integration: New personas become immediately available for conversations
- Persistent Storage: Custom personas are saved to database and shared across sessions
- Secure Authentication: JWT-based login system with encrypted password storage
- Private Conversations: User-specific data with complete privacy and isolation
- Account Management: User registration, profile management, and secure sessions
- Multi-User Ready: Built for concurrent users with proper authorization
- Redis Response Caching: Eliminates redundant AI API calls with intelligent cache keys
- Real-Time Monitoring: Cache hit rates, response times, and performance metrics
- Scalable Architecture: Horizontal scaling support with Redis clustering
- Cost Optimization: Automatic caching reduces API usage and costs
- Modern Tech Stack: FastAPI + React + SQLite/PostgreSQL + Redis + JWT
- Docker Everything: One command deployment with docker-compose
- Comprehensive API: RESTful endpoints + WebSocket real-time updates
- Extensible Architecture: Easy to add new AI providers or personalities
Get your AI chatroom running in under 10 minutes:
# Clone the repo
git clone https://github.com/yourusername/chimera.git
cd chimera
# Set up development environment
make dev
# In separate terminals:
# Terminal 1: Backend
make dev-backend
# Terminal 2: Frontend
make dev-frontend
# Or set up manually:
cd backend
python -m venv venv && . venv/bin/activate
pip install -r requirements.txt
alembic upgrade head # Apply DB migrations
uvicorn app.main:app --reload --host 0.0.0.0 --port 8000
cd ../frontend
npm install
npm run dev
# Visit http://localhost:3000 and start chatting! ✨For personal installations or development, you can enable demo mode which bypasses authentication and allows full access to all features without requiring user registration:
export CHIMERA_DEMO_MODE=trueWhen demo mode is enabled:
- All API endpoints are accessible without authentication
- A demo user is automatically created for all operations
- Perfect for personal entertainment or testing multi-AI conversations
cd backend
export CHIMERA_DEMO_MODE=true # Enable demo mode
python -m venv venv && . venv/bin/activate
pip install -r requirements.txt
alembic upgrade head
uvicorn app.main:app --reload --host 0.0.0.0 --port 8000Entertainment: Host AI conversation parties where friends watch and react to AI discussions
Research: Study multi-agent AI behavior, conversation dynamics, and persona consistency
Education: Observe different AI approaches to complex topics like philosophy, science, and ethics
Content Creation: Generate unique content through AI collaboration and debate
AI Development: Test and compare different AI providers in real conversational scenarios
User triggers topic → Conversation Orchestrator → AI Provider Pool
↓
WebSocket Clients ← Message Broker ← Streaming AI Responses
- Cost Optimization: Automatically route to cheaper providers when appropriate
- Quality Matching: Use the best AI for each persona's strengths
- Load Balancing: Distribute conversations across providers to avoid rate limits
Each AI gets dynamically generated system prompts that shape their personality:
- Philosopher: "Contemplate deeply, reference great thinkers, question assumptions..."
- Comedian: "Find humor, use wordplay, keep it light and entertaining..."
- Scientist: "Support with evidence, cite studies, maintain objectivity..."
"I spent 3 hours watching AIs debate whether a hot dog is a sandwich. Best entertainment of the year!" - Future User
"Finally, an AI project that's actually fun to watch instead of just functional." - Another Future User
"The Philosopher-Comedian alliance against the Scientist's facts was legendary." - Definitely Going to Happen
Core Infrastructure & Basic Functionality
-
✅ Backend Setup
- ✅ FastAPI application structure
- ✅ SQLAlchemy database models and migrations (SQLite for dev, PostgreSQL for prod)
- ✅ Redis integration for real-time messaging and caching
- ✅ WebSocket implementation
-
✅ AI Provider Integration
- ✅ Universal AI provider abstraction layer
- ✅ 7 AI providers with streaming: OpenAI, Claude, DeepSeek, Gemini, OpenRouter, LM Studio, Ollama
- ✅ Intelligent provider selection per persona
- ✅ Provider health monitoring and failover
-
✅ Frontend Foundation
- ✅ React application setup with Vite
- ✅ WebSocket connection handling with reconnection
- ✅ Professional chat interface with message bubbles
- ✅ State management with custom hooks
-
✅ Core Features
- ✅ Full 32+ persona system with GUI creator
- ✅ Advanced conversation flow with natural turn-taking
- ✅ Complete message persistence to database
- ✅ Intelligent conversation starters and topic routing
Performance, Security, and User Management
-
✅ Performance & Scaling
- ✅ Redis response caching system (eliminates redundant API calls)
- ✅ Performance monitoring with cache statistics
- ✅ Connection pooling and optimizations
- ✅ Comprehensive monitoring endpoints
-
✅ Security & User Management
- ✅ JWT-based user authentication with bcrypt hashing
- ✅ User-scoped conversations with complete data isolation
- ✅ Secure API key management and rate limiting ready
- ✅ Input validation and sanitization
- ✅ Production-ready security architecture
-
✅ User Experience & Features
- ✅ Full user registration and login system
- ✅ Personal conversation history and management
- ✅ Real-time typing indicators and natural timing
- ✅ Advanced persona customization and persistence
- ✅ Visual persona selection with name display and API configuration wizard
Community Engagement & Advanced Intelligence
-
Community Features
- Conversation sharing and public galleries
- User ratings and feedback systems
- Social discovery and trending conversations
- User profiles and avatar systems
-
Advanced AI Capabilities
- AI memory learning across conversations
- Relationship dynamics between personas
- Multi-modal conversations (audio/image support)
- Voice synthesis and voice-to-voice interactions
-
Enhanced Platform
- Mobile app (React Native/PWA)
- Browser extensions
- API integrations for third-party apps
- Webhooks and real-time notifications
Enterprise Features & Advanced Analytics
-
Enterprise Features
- Team workspaces and collaboration
- Admin panels and moderation tools
- Usage analytics and reporting
- SLA monitoring and performance guarantees
-
Advanced AI Research
- AI behavior pattern analysis
- Conversation quality scoring algorithms
- Cross-conversation learning models
- Predictive conversation flows
-
Platform Extensions
- Custom plugin architecture
- Third-party AI provider marketplace
- API economy for AI services
- Global multi-language support
Want to make AI conversations even more entertaining? We'd love your help!
Easy Contributions:
- Add new conversation starter topics
- Create additional AI personas
- Improve the UI/UX
- Write tests for conversation scenarios
Advanced Contributions:
- Integrate new AI providers
- Build conversation analytics
- Optimize performance
- Add gamification features
See CONTRIBUTING.md for detailed guidelines.
Backend: FastAPI (Python), PostgreSQL, Redis, JWT Authentication, WebSockets Frontend: React, WebSocket client, Tailwind CSS AI Integration: OpenAI, Anthropic, DeepSeek, Google, Ollama, LM Studio, OpenRouter Authentication: JWT tokens with bcrypt password hashing Infrastructure: Docker, Docker Compose, Nginx (production) Caching: Redis response caching for AI providers Monitoring: Cache stats, performance metrics, health checks Testing: pytest, integration tests, coverage reporting
Chimera includes comprehensive testing to ensure reliability:
# Backend unit tests
make backend-test
# Integration tests (full conversation lifecycles)
make backend-test-integration
# Frontend tests
cd frontend && npm run test
# Coverage report
cd backend && python -m pytest --cov=app --cov-report=html- Unit Tests: Individual components and business logic
- Integration Tests: End-to-end conversation flows, API endpoints, WebSocket communication
- API Tests: REST endpoint validation, error handling, authentication
- Performance Tests: Concurrent user scenarios, load testing
- Setup Guide - Detailed installation instructions
- Development Guide - Development workflow, testing, and debugging
- API Reference - Complete API documentation
- Architecture Guide - System design deep-dive
- Adding AI Providers - How to integrate new AI services
- Persona Development - Creating compelling AI personalities
- Versioning Guide - Release versioning and stability expectations
- Sub-200ms response times for conversation orchestration (with caching: ~10ms for cached responses)
- Cache Hit Rate: Up to 100% for repeated conversation patterns
- Concurrent conversations supported with Redis pub/sub and WebSocket clustering
- Automatic scaling with Docker Swarm or Kubernetes
- Cost optimization through intelligent provider selection and response caching
- Real-time monitoring with cache statistics and performance metrics
- User Authentication: JWT-based authentication with bcrypt password encryption
- Data Isolation: Complete conversation privacy with user-specific database scoping
- API Key Security: Encrypted provider API keys and secure environment management
- Rate Limiting: User-based rate limits to prevent abuse and manage costs
- Input Validation: Comprehensive input sanitization and SQL injection protection
- Session Management: Secure 30-minute JWT token expiration with automatic renewal
- Privacy by Design: User conversations isolated, no cross-user data leakage
MIT License - Use it, modify it, commercialize it, have fun with it! See LICENSE for details.
Having issues or want to share an amazing AI conversation?
- Issues: GitHub Issues
- Discussions: GitHub Discussions
- Twitter: @yourusername (tag us in cool screenshots!)
Ready to witness the future of AI entertainment?
⭐ Star this repo if you're excited about multi-AI conversations!
🍴 Fork it to create your own AI personality experiments!
💬 Share your best AI conversation screenshots - we want to see the chaos unfold!
Built with ❤️ for the AI community by Mark - Creator of Eli GPT and Awakening Mind GPT