LocalSeek 🤖💬

LocalSeek is a powerful, privacy-first AI chat extension for Visual Studio Code that brings conversational AI directly to your development environment - completely locally.

🌟 Key Features

💻 Fully Local AI Interaction

• Zero Cloud Dependencies: All AI processing happens 100% on your machine
• Privacy-First Approach: No external API calls or data transmission
• Offline-Ready: Works without internet connection
• Local Model Management: Seamlessly use any Ollama-compatible model

🌏 Visually Stunning

• NeverBeforeSeen Look: Beautiful UI with Webview API integrated directly within VSCode
• Markdown Support: Supports MD on the STREAMED AI's response, so it is easy to understand
• Talk, Copy, Exit: User friendly chat interface like mobile phones with many features

🚀 Seamless VSCode Integration

• Integrated sidebar AI chat view
• Standalone chat panel
• Instant model switching
• Lightweight and non-intrusive design

🔒 Local Processing Advantages

• Enhanced Privacy: Your conversations never leave your computer
• Reduced Latency: Direct local model access
• Cost-Effective: No subscription or per-token charges
• Customizable: Use any locally hosted AI model

📋 Prerequisites

System Requirements

• Visual Studio Code (v1.96.0+)
• Ollama installed locally
• Minimum 8GB RAM recommended
• At least one Ollama-compatible LLM model (DeepSeek R1 is preffered 😉)

Recommended Model Installations

# Pull recommended models
ollama pull deepseek-r1:14b   # The best in my opinion
ollama pull mistral           # Balanced performance
ollama pull llama3.2          # Versatile model
ollama pull phi3              # Lightweight option

🔧 Installation

Method 1: Visual Studio Code Marketplace

1. Open VSCode Extensions (Ctrl+Shift+X)
2. Search for "LocalSeek"
3. Click "Install"

Method 2: Manual VSIX Installation

1. Download VSIX from Releases
2. Open VSCode Extensions view
3. Click "..." menu
4. Select "Install from VSIX"
5. Choose downloaded file

🚀 Quick Start Guide

Accessing LocalSeek

• Sidebar Chat: Click LocalSeek icon in Activity Bar 👈
• Standalone Chat:
  • Command Palette (Ctrl+Shift+P)
  • Type "LocalSeek AI Chat"

Basic Usage

1. Select Ollama model from dropdown
2. Type your message
3. Press Enter or Send button
4. Receive instant, local AI responses

⚙️ Configuration

Ollama Host Configuration

Customize Ollama connection in VSCode Settings:

• Open Settings (Ctrl+,)
• Search "LocalSeek"
• Modify "Ollama Host" if needed

🔬 Supported Models

LocalSeek works with any Ollama-compatible model including but not limited to:

• DeepSeek-R1
• Mistral
• Llama 3
• Phi-3
• CodeLlama
• Mixtral
• ...

🛡️ Privacy Commitment

• 100% Local Processing
• No External Data Transmission
• Full Control Over Your AI Interactions

🤝 Contributing

Contributions welcome!

• Report issues on GitHub
• Submit pull requests
• Suggest new features

📦 Troubleshooting

• Ensure Ollama is running
• Verify model installations
• Check VSCode and extension compatibility

📝 License

MIT License

Developed with ❤️ by Hariharen