NativeMind
Private, On-device Al Assistant

NativeMind is a private, open-source alternative to ChatGPT — a fully local browser extension built on Ollama that brings local AI models into your workflow.

---

The Future of Private AI is Here, Running Entirely on Your Device

NativeMind delivers cutting-edge intelligence without cloud dependencies—faster responses, zero tracking, unlimited usage—all running quietly in your browser while you stay in complete control.

⭐ Getting Started & Stay Updated

Star this project to get all the latest updates — you'll never miss a new release!

✨ Core Features

🤖 AI Agent - Your Intelligent Assistant

• Autonomous Task Execution: AI agents that can understand complex requests and break them down into actionable steps
• Multi-Step Problem Solving: Handle complex workflows that require multiple tools and reasoning steps
• Context-Aware Decision Making: Agents understand your current context and make intelligent decisions
• Tool Integration: Seamlessly use various tools and capabilities to accomplish your goals

💬 Browse Smarter with Local AI

• Contextual Awareness Across Tabs: Seamlessly continue AI conversations with memory that spans across tabs

• AI-Powered Local Search: Ask questions and search the web directly from your browser—no third-party APIs required

• Smart Page Understanding: Instantly summarize and interpret any webpage content

• Bilingual Translation: Translate full pages or selected text with side-by-side view (supports multiple languages)

• Quick Actions: Custom Skills for any task, adapt NativeMind to match your workflow with local model capabilities

  

✍️ Boost Productivity with AI Tools

• Chat with PDFs: Ask questions directly about PDF files with intelligent text extraction

• Chat with Images: Understand and interact with visual content, including screenshots and uploaded images

  

• Advanced Writing Tools: Comprehensive writing assistance including rewriting, proofreading, tone adjustment, and creative rephrasing

  

🧠 Run Your Own AI Models Locally

• Offline-First: Use all features without an internet connection after downloading a model—your data stays on your device
• Model Switching: Easily switch between models based on performance, speed, or privacy needs

📦 Installation

NativeMind supports Chrome, Firefox, Brave, and Edge browsers with full feature compatibility across all platforms.

🚀 Quick Start

⚡ First-Time Setup

For more details, visit our official website: nativemind.app

1. Pin the Extension: Click the extension icon and pin it to your toolbar
2. Setup Ollama: Follow the Ollama Setup Guide for full AI capabilities
3. Start Using: Click the NativeMind icon in your browser toolbar

🤖 AI Model Options

🦙 Ollama (Recommended)

• Advanced Models: Deepseek, Qwen, Llama, Gemma, Mistral, Phi and more

• gpt-oss Support: Full compatibility with gpt-oss models for enhanced open-source AI capabilities

  

• Superior Performance: Leverage full system resources

• Model Flexibility: Download and switch between specialized models

• Local Server: Self-hosted for maximum control

⚡ WebLLM (Quick Trial)

• Instant Demo: Try AI capabilities immediately with Qwen3-0.6B in your browser
• Zero Setup: No additional software installation required for quick testing
• WebAssembly-Powered: Runs directly in browser using WebAssembly
• Trial Experience: Perfect for evaluating the extension before setting up Ollama

🔮 Why On-Device AI？

⚠️ The Problem with Cloud AI

In an era where every keystroke can be monitored and every document potentially analyzed, cloud-based AI services represent a fundamental privacy risk. Your sensitive data—personal documents, business communications, creative work—becomes part of someone else's training data or surveillance apparatus.

🚀 Performance Revolution

Privacy no longer means compromise. Recent breakthroughs in model efficiency mean on-device AI now outperforms cloud alternatives:

• Qwen3-4B: Outperforms Qwen2.5-72B while being 18x smaller
• Gemma3-4B: Matches Gemma2-27B performance with 6.75x fewer parameters and exceptional image recognition capabilities
• Phi-4: Outperforms Gemini Pro 1.5 in mathematical reasoning and competition problems
• Zero Latency: No network delays mean instant responses
• Always Available: No API limits, rate limiting, or service outages

🔒 True Data Sovereignty

NativeMind puts you back in control. By running AI models entirely on your device, we eliminate the fundamental privacy risks of cloud AI:

• 100% On-Device Processing: Every keystroke, every document, every conversation stays on your device
• Zero Cloud Dependencies: No data ever transmitted to external servers
• No Corporate Surveillance: No tech giant can analyze your usage patterns or content
• No Government Access: No backdoors or potential government data requests
• No Data Breaches: If your data never leaves your device, it can't be stolen from servers
• No Account Required: Use all features without registration or login
• Open Source Transparency: Every line of code is publicly auditable

🧩 How Does NativeMind Compare?

Wondering how NativeMind compares to ChatGPT or Ollama? Here’s a quick breakdown to help you see where it stands — and when it might be the better choice.

🛠️ For Developers

Technology Stack

Prerequisites

• Node.js (v22.14.0 recommended)
• PNPM (v10.10.0 or later)
• Modern browser (Chrome, Firefox, or Edge)

Installation

1. Clone the repository

git clone https://github.com/NativeMindBrowser/NativeMindExtension.git
cd NativeMindExtension

2. Install dependencies

pnpm install

3. Start development server

pnpm dev         # Chrome development (default)
pnpm dev:edge    # Edge development

Building for Production

# Development build
pnpm zip:beta

# Production build  
pnpm zip:prod

The built extension will be in the .output/ directory.

Manual Installation (Development)

1. Download: Get the latest release from GitHub Releases
2. Install: Open chrome://extensions/ → Enable "Developer mode" → Click "Load unpacked"
3. Setup: Pin the extension and choose your AI model

🔨 Troubleshooting

Common Issues

Extension Not Loading

• Ensure Developer mode is enabled in browser extensions page
• Check that all files were extracted/loaded correctly
• Try refreshing the extensions page and reloading

AI Models Not Working

• WebLLM: Ensure sufficient RAM (8GB+ recommended) and modern browser
• Ollama: Verify Ollama is running locally on port 11434
• Clear browser cache and restart extension
• Check browser console for error messages

Slow Performance

• WebLLM: Close other tabs to free up memory
• Ollama: Ensure sufficient system resources
• Try smaller AI models for faster response times
• Check if antivirus software is interfering

Translation/Summarization Not Working

• Verify AI model is properly loaded
• Check if page has restricted content policies
• Try refreshing the page and retrying
• Ensure page content is selectable text (not images)

Getting Help

• Check browser console (F12 → Console) for error messages
• Join our Discord community for support: https://discord.gg/b8p54DKhha
• Report issues on GitHub Issues
• Include browser version, OS, and error details in reports

🤝 Contributing

🚀 Quick Start for Contributors

1. Fork the project
2. Create your feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add some amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

📋 Development Guidelines

• ✅ Follow the existing code style
• ✅ Add tests for new functionality
• ✅ Update documentation as needed
• ✅ Ensure privacy-first principles

📄 License

🙏 Acknowledgments

• 👥 Contributors: Thanks to all who have helped shape NativeMind
• 🛠️ WXT: Making browser extension development easier
• 🧠 WebLLM: Enabling in-browser AI capabilities
• 🦙 Ollama: Powering local LLM support