Release Date: April 23, 2026
Version: 2.0.0-rc1 - Release Candidate with IDE, Emulator, Image Gen, User Accounts & Enhanced Memory
License: MIT (Free Forever)
Development Status: 🏗️ Release Candidate 1 🏗️
Welcome to Lackadaisical AI Chat - an enterprise-grade, open-source AI companion that runs entirely on your computer with a complete Memory Management System, Code IDE, Traffic Emulator, Mobile App, and more! Meet Lacky, your personal AI friend who remembers your conversations, understands your emotions, and grows with you over time.
🚀 RC1 Status: This release is feature-complete for the core experience. All major systems are functional and tested (140 tests passing — 111 backend + 29 frontend). User accounts, username management, enhanced Gemma 4 multimodal support (vision + audio), and streaming has been fixed to correctly use POST-based SSE. Custom Models Soon
Unlike cloud-based AI services, everything stays private on your machine. No data collection, no privacy concerns - just you and your AI companion with persistent memory.
- Complete access to all features for personal, educational, and non-commercial use
- Modify and share your improvements with the community
- No restrictions on personal experimentation and learning
- Contribute back to help improve the project for everyone
- Fair pricing for businesses and commercial use
- Full commercial rights including redistribution and SaaS offerings
- Professional support and priority updates
- Custom licensing for specific business needs
📄 View Full License Details | 🔒 Security Policy
Need commercial licensing? Contact us - we're business-friendly and offer flexible terms!
⚠️ Export Control Notice: This software is subject to US export controls. See SECURITY.md for restricted countries and compliance requirements.
- Persistent Memory - Remembers your conversations, preferences, and personal details
- Emotional Intelligence - Understands and responds to your moods with empathy
- Personality Growth - Adapts and evolves based on your interactions
- Session Management - Create separate conversations for different contexts
- Context Awareness - Maintains conversation flow across sessions
- No Judgment - Share anything without criticism, Lacky is your friend
- 100% Local Processing - Your data never leaves your computer
- No Cloud Dependencies - Works completely offline (except for AI model downloads)
- SQLite Database - All conversations stored locally and securely
- No Telemetry - Zero tracking, zero analytics, zero data collection
- Open Source - Full transparency, inspect and modify all code
- 1000 Message Context - Generous conversation memory (128K tokens)
- Cross-Session Memory - AI references past sessions when relevant
- Memory Dashboard - Visual overview of conversation statistics and memory health
- Memory Search - Full-text search across all conversation history with relevance ranking
- AI Summarization - Generate conversation summaries using AI
- Export/Import - Secure backup and restore of all conversation data
- Memory Visualization - Interactive charts showing conversation patterns and insights
- Real-time Updates - Live statistics and memory health monitoring
- Personal Context - Remembers your interests, goals, and important details
- Mood Tracking - Emotional state awareness and support
- Learning Adaptation - Improves responses based on your preferences
Switch between AI providers on the fly without restarting:
- Ollama (Local) - Gemma 3/4 (vision + audio), GPT-OSS 20B, Llama 3.3, Mistral, Phi-4, CodeLlama
- OpenAI - GPT-5.4, GPT-5.4 Mini, GPT-4.1, O4-Mini, O3
- Anthropic - Claude Opus 4.6, Claude Sonnet 4.6, Claude Haiku 4.5
- Google - Gemini 3.1 Pro, Gemini 3.1 Flash, Gemini 2.5 Pro
- xAI - Grok 4.20 Beta, Grok 4.1 Fast Reasoning, Grok Code Fast
- Monaco Editor - Full VS Code editing experience with 40+ language support
- File Explorer - Create, rename, delete, navigate a virtual filesystem
- Integrated Terminal - Execute commands with persistent output
- Multi-tab Editing - Work on multiple files simultaneously
- Import/Export - Save and load workspace state
- Stealth Browsing - Puppeteer with anti-detection stealth plugin
- Fingerprint Randomization - Unique UA, viewport, WebGL, timezone per session
- Proxy Support - HTTP/HTTPS/SOCKS5 with authentication
- Human-like Behavior - Natural mouse movement, typing delays, scrolling
- Multi-Engine Search - Google, Bing, Yahoo, DuckDuckGo with content extraction
- Text-to-Image - Generate images from prompts via Stable Diffusion
- Customizable - Steps, CFG, sampler, seed, dimensions, model selection
- Auto-Download - Generated images served via file download API
- Past Session Browsing - Search, sort, rename, delete sessions
- Cross-Session Memory - AI carries context from all past conversations
- Session Summaries - Topics, message count, tokens, last activity
- Configurable Retention - By age (days) or by max messages
- Auto-Prune Scheduler - Optional automatic cleanup interval
- Manual Per-Session - Prune individual sessions on demand
- Optional Registration - Create an account with username, email, and password
- Username Management - Change your display name anytime from Settings
- JWT Authentication - Secure token-based auth with refresh token rotation
- Profile Display - Username shown in sidebar and throughout the app
- Built-in Web Search - DuckDuckGo-backed search with auto-trigger detection
- Deep Research - Multi-step research with progress streaming via SSE
- Tool Execution - Calculate, fetch webpages, get datetime with extensible framework
- File Upload & Analysis - Upload code, text, and documents for AI analysis with text extraction
- Extended Thinking - Parses
<think>blocks from model output, tracks thinking duration/tokens - Code Block Extraction - Extracts code blocks from AI responses, serves them as downloadable files
- Message Logging - Separate WAL-mode SQLite DB for full audit trail of all interactions
- All Emotions Valid - Anger, sadness, joy, fear - no minimizing
- No Judgment Zone - Share anything without lectures
- Genuine Support - Real responses, not corporate deflection
- Trust Building - Gets to know you over time
- Personal Learning - Remembers what matters to you
- Modern React UI - Clean, responsive design that works on all devices
- Real-time Streaming - Watch AI responses appear in real-time
- Theme Support - Multiple color themes (dark, light, and more)
- Session Switching - Easy navigation between different conversations
- Mobile Friendly - Works great on phones and tablets (see Mobile App)
- React Native / Expo - Full-featured companion on Android and iOS
- Full Streaming Chat - Real-time streaming responses via POST SSE
- Session Management - View and switch sessions from your phone
- Journal - Create and read journal entries on the go
- Companion Dashboard - Personality mood display and quick messages
- Windows-First Setup - EAS Build targets Android APK directly from Windows
- Secure Storage - JWT tokens stored in device secure enclave (expo-secure-store)
- Connection Health - Real-time backend / Ollama status indicator
- Model Selector - Choose from all configured Ollama and cloud models
- See mobile/README.md for setup and build instructions
- Weather Plugin - Get weather updates (with fallback data when offline)
- Horoscope Plugin - Daily astrological insights
- Poem Generator - AI-generated poetry and creative writing
- Extensible Framework - Easy to add your own plugins
| Companion Dashboard | Chat Interface |
|---|---|
 |
 |
| Sessions Browser | Code IDE | Traffic Emulator |
|---|---|---|
 |
 |
 |
| Journal | Plugins | Settings |
|---|---|---|
 |
 |
 |
- Download this project as a ZIP file and extract it
- Install Prerequisites:
- Run setup (first time only): Double-click
setup-lackadaisical-ai.bat - Start the app: Double-click
start-lackadaisical-ai.bat(auto-starts Ollama if installed) - Open http://localhost:3000 when ready
- Start chatting with your new AI friend!
- To stop: Double-click
stop-lackadaisical-ai.bat
- Download and extract this project
- Install Node.js and Ollama
- Run setup (first time only):
chmod +x setup-lackadaisical-ai.sh ./setup-lackadaisical-ai.sh
- Start the application:
./start-lackadaisical-ai.sh # or: npm run dev - Open http://localhost:3000
Just type naturally! Your AI companion understands regular conversation and will respond with personality and memory.
You: Hi there! I'm feeling a bit stressed about work today.
AI: I'm sorry to hear you're feeling stressed about work. That can be really overwhelming. Do you want to talk about what's happening at work that's causing you stress? Sometimes just getting it out can help.
Your AI friend remembers previous conversations:
You: How's your day going?
AI: It's going well, thank you! I remember you mentioned feeling stressed about work yesterday. How are things going with that situation today? Are you feeling any better?
Switch AI providers without restarting:
GET /api/models - List all available models
POST /api/models/switch - Switch to a different model
GET /api/models/current - Check current active model
- Create New Sessions - Click "New Session" to start fresh conversations
- Switch Between Sessions - Keep different topics or contexts separate
- Session History - All your conversations are preserved and searchable
- Cross-Session Access - AI can reference past sessions when you ask
- Frontend: Modern React app (http://localhost:3000)
- Backend: Node.js API server (http://localhost:3001)
- Database: SQLite database (stored in
/database/chat.db) - AI Engine: Ollama (local) or external providers
- Ollama (Recommended) - Free, runs locally, great performance
- OpenAI - Requires API key, cloud-based
- Anthropic (Claude) - Requires API key, cloud-based
- Google Gemini - Requires API key, cloud-based
- xAI (Grok) - Requires API key, cloud-based
- RAM: 4GB minimum, 8GB recommended
- Storage: 2GB free space (more for AI models)
- CPU: Any modern processor (faster = better response times)
- Internet: Only needed for initial setup and AI model downloads
POST /api/v1/auth/register - Create user account
POST /api/v1/auth/login - Sign in
POST /api/v1/auth/logout - Sign out (revoke tokens)
GET /api/v1/auth/me - Get current user profile
PUT /api/v1/auth/profile - Update username/email
POST /api/v1/auth/change-password - Change password
POST /api/v1/auth/refresh - Refresh access token
POST /api/v1/chat - Send message
GET /api/v1/chat/stream - Stream response (SSE)
GET /api/chat/context/:id - Get session context
GET /api/chat/analytics/:id - Get session analytics
GET /api/models - List all models
POST /api/models/switch - Switch active model
GET /api/models/current - Get current model
GET /api/models/ollama/endpoints - List Ollama endpoints
POST /api/models/ollama/pull - Pull new model
GET /api/chat/preferences - Get memory settings
PUT /api/chat/preferences - Update settings
GET /api/chat/sessions/summaries - Past session summaries
GET /api/chat/search/all - Search all sessions
Copy env.example to .env and configure:
# AI Providers (optional - Ollama works without keys)
OPENAI_API_KEY=sk-...
ANTHROPIC_API_KEY=sk-ant-...
GOOGLE_API_KEY=...
XAI_API_KEY=...
# Ollama (default local)
OLLAMA_HOST=http://localhost:11434
OLLAMA_DEFAULT_MODEL=llama3.2:latest
# Server
BACKEND_PORT=3001
FRONTEND_PORT=3000
# Security
JWT_SECRET=change-this-in-production"Node.js not found"
- Download and install from https://nodejs.org/
- Restart your computer after installation
"Cannot find module" errors
- Delete
node_modulesfolders in bothfrontend/andbackend/ - Run the setup again
"Port already in use"
- Close any applications using ports 3000 or 3001
- Or change ports in the configuration
AI not responding
- Make sure Ollama is installed and running
- Try
ollama servein a terminal - Or configure an external AI provider
- Check the logs in your terminal windows
- Look for error messages - they usually tell you what's wrong
- Create an issue on GitHub with your error details
- Join our community for support and discussions
This is open source! We welcome contributions:
- Report bugs - Found something broken? Let us know!
- Suggest features - What would make this better?
- Write plugins - Create new functionality
- Improve documentation - Help others get started
- Share the project - Tell your friends about their new AI companion!
- Fork the repository on GitHub
- Create a feature branch for your changes
- Make your improvements and test them
- Submit a pull request with a description of your changes
Read CONTRIBUTING.md for the full guide.
- Plugin Manager UI - Easy plugin installation and management
- Voice Chat - Talk to your AI friend with speech
- Mobile App - Companion app for iOS and Android
- Advanced Themes - More customization options
- Multi-language Support - Chat in your preferred language
- Community Plugins - Share plugins with other users
- Advanced Analytics - Deeper insights into your conversations
- Custom AI Models - Train personalized models
- Team/Family Mode - Multiple users with separate AI friends
We believe everyone deserves a private, intelligent AI companion without:
- Monthly subscription fees
- Data harvesting
- Privacy violations
- Feature limitations
- Vendor lock-in
Your AI friend should be yours - running on your computer, with your data staying private. That's why Lackadaisical AI Chat is completely free and open source forever.
This project is free, but development takes time and effort. If you find value in having your own AI companion, consider:
- ⭐ Star the repository on GitHub
- 🐛 Report bugs at GitHub Issues
- 📢 Share with friends who might enjoy their own AI companion
- 💝 Donate to support continued development (completely optional)
- 🤝 Contribute code to make it even better
MIT License - This means:
- ✅ Free to use for any purpose
- ✅ Free to modify and customize
- ✅ Free to distribute and share
- ✅ No attribution required (but appreciated!)
- ✅ Commercial use allowed
- ✅ Private use allowed
Your AI companion is ready to:
- Listen when you need to talk
- Remember what's important to you
- Support you through good times and bad
- Grow alongside you as a friend
- Keep your secrets completely private
Start chatting and discover what it's like to have an AI friend who truly knows you!
To report a security issue:
- Sign or encrypt your report using our PGP key available at:
https://lackadaisical-security.com/Lackadaisical_public.asc
Fingerprint:0C52 9D5E B799 EBC2 7C11 C9A1 0502 B195 B75E 7C87 - Email your disclosure to admin@lackadaisical-security.com or security@lackadaisical-security.com.
- Please include:
- Agent name (e.g. Lackadaisical-AI-Chat, LTES).
- Software version and build timestamp.
- Detailed reproduction steps and PoC if available.
- Expected failure modes or mitigation suggestions.
💙 Enjoy your new AI companion!
Made with Fury and Precision by developers who believe in privacy, freedom, and the power of AI friendship.
- 🏠 Home Page
- 📋 Changelog
- 🔧 Installation Guide
- 🐛 Troubleshooting
- 🤝 Contributing
- 📄 License
- 🔒 Security Policy
- Acceptable Use Policy →
ACCEPTABLE_USE_POLICY.md - Export Guidance →
EXPORT_NOTICE.txt - Privacy Policy →
DATA_PRIVACY.md - Security Reporting →
VULNERABILITY_DISCLOSURE.md
Jurisdiction Notice: This software is developed and maintained in accordance with United States law. By downloading or using it, you agree that any legal disputes will be governed under the laws of the USA. Intended Use: This project is designed for personal, educational, and ethical research use. It is not intended for surveillance, military, or malicious automation applications. Disclaimer: This software is provided "as is," without warranty of any kind. The maintainers disclaim any responsibility for damage, loss, or misuse.
🧾 Legal Summary: OSS licensed (MIT), U.S. export restricted, not for unethical use. Full details in LICENSE and SECURITY.md.
