🤖 O10D: LocalAI | Offline AI Chat & Coding Assistant

Privacy-first AI coding assistant that runs completely offline with local LLMs 🚀

O10D: LocalAI is a powerful VS Code extension that brings AI-powered coding assistance directly to your editor without compromising your privacy. Everything runs locally on your machine! 🏠✨

Cross-Platform Support: Fully compatible with Windows, macOS, and Linux (including Ubuntu and Fedora). 🐧🍎🪟

💡 Tip: Move to Right Side (Secondary Sidebar)

Right-click the LocalAI icon in the Activity Bar
Select "Move to Secondary Side Bar"
Now LocalAI will appear on the right side! ➡️ OR simply drag and drop the LocalAI view to the right sidebar

💡 Be patient - First result may take a moment as the LLM has to load into memory and process the context

✨ Features

🎯 Core Functionality

💬 Smart Chat Interface - Interactive chat panel with your AI coding assistant
📝 Code Explanation - Get detailed explanations of your selected code
🔧 Code Refactoring - Improve code readability and performance with AI suggestions
🔍 Semantic Code Search - Intelligent indexing and search across your entire codebase
📁 Context Management - Add files (including CSV & Excel spreadsheets) and folders to provide context for better responses
💾 Persistent Chat History - Your conversations are saved per workspace
🔄 Tab Chat Mode - Open chat in a dedicated tab for focused conversations

🎨 User Experience

🎭 UI - Modern, clean interface with smooth animations
⚡ Real-time Streaming - See responses as they're generated
🛑 Stop Response - Cancel long-running responses anytime
📋 Copy Responses - Easy one-click copying of AI responses
🎯 Smart Context - Automatically includes relevant file context
📱 Responsive Design - Works perfectly in both sidebar and tab modes
🗑️ Model Management - Easily view and uninstall installed AI models directly from the UI

💻 System Requirements

Because O10D: LocalAI runs Large Language Models (LLMs) locally on your machine, your hardware specifications directly impact performance and response times.

Hardware Specifications

Component	Minimum	Recommended
CPU	Intel Core i5 / AMD Ryzen 5 (or equivalent)	Apple Silicon (M1/M2/M3/M4) or modern multi-core CPU
GPU	Integrated graphics / CPU-only execution	Dedicated NVIDIA GPU (6GB+ VRAM) or Apple Silicon Unified Memory
RAM	8 GB RAM (suitable for `gemma3:270m` / `gemma3:4b`)	16 GB RAM or more (suitable for `deepseek-coder:6.7b` / `qwen2.5-coder:7b` / `llama3.1:8b`)
Storage	10 GB free space (SSD preferred)	20 GB+ free space on a fast NVMe SSD

Software Requirements

VS Code: Version 1.70.0 or higher
OS: Windows 10/11, macOS 12+, or modern Linux distributions (Ubuntu, Fedora, etc.)

🚀 Getting Started

📥 Installation

Install the Extension 📦
- Open VS Code
- Go to Extensions (Ctrl+Shift+X / Cmd+Shift+X)
- Search for "O10D: LocalAI"
- Click Install
First Launch 🎉
- The extension will automatically guide you through setup
- Follow the prompts to get started

🎮 How to Access

Look for the 🤖 robot icon in the Activity Bar (left sidebar)
Click to open the LocalAI chat panel

Method 2: Command Palette ⌨️

Press Ctrl+Shift+P (Windows/Linux) or Cmd+Shift+P (Mac)
Type "LocalAI" to see all available commands

🎮 Commands

Toggle Auto-Suggestions: Ctrl+Shift+P → "LocalAI: Toggle Auto-Suggestions"
Clear Suggestion Cache: Ctrl+Shift+P → "LocalAI: Clear Suggestion Cache"

💡 Tips for Best Results

Provide context - The more context around your cursor, the better the suggestions
Be patient - First suggestions may take a moment as the LLM processes the context

Method 3: Keyboard Shortcuts ⚡

Ctrl+Shift+L (Windows/Linux) or Cmd+Shift+L (Mac) - Open Chat
Ctrl+Shift+E (Windows/Linux) or Cmd+Shift+E (Mac) - Explain Selection
Ctrl+Shift+R (Windows/Linux) or Cmd+Shift+R (Mac) - Refactor Selection

🛠️ Available Commands

💬 Chat Commands

Command	Description	Shortcut
`LocalAI: Open Chat`	Open the main chat interface	`Ctrl+Shift+L` / `Cmd+Shift+L`
`LocalAI: Clear Chat History`	Clear all chat history for current workspace	-

🎯 Code Analysis & Agent Commands

Command	Description	Shortcut
`LocalAI: Explain Selection`	Get detailed explanation of selected code	`Ctrl+Shift+E` / `Cmd+Shift+E`
`LocalAI: Refactor Selection`	Get refactored version of selected code	`Ctrl+Shift+R` / `Cmd+Shift+R`
`LocalAI: Generate Tests`	Generate unit tests for selected code	-
`LocalAI: Toggle Agent Mode`	Enable or disable autonomous agent operations	-
`LocalAI: Semantic Code Search`	Perform intelligent codebase search	-
`LocalAI: Reindex Workspace`	Rebuild the semantic search index	-

📁 Context & Server Management

Command	Description
`LocalAI: Add Files/Folders as Context`	Add files or folders to chat context
`LocalAI: Check Status`	Check extension and server status
`LocalAI: Restart Server`	Restart the local server

🎨 Interface Guide

💬 Chat Panel

📝 Input Area - Type your questions or requests
📎 Context Button - Add files/folders to provide context
🗑️ Clear Button - Clear chat history
📑 Tab Button - Open chat in a dedicated tab
📤 Send Button - Send your message
⚙️ More Menu - Access additional options like Uninstall Model

🎯 Context Files

📁 File Picker - Browse and select files/folders
📊 Spreadsheet Support - Seamlessly process and extract context from CSV and Excel files
🏷️ File Chips - Visual indicators of added context files
❌ Remove Context - Click the X to remove files from context

⚙️ Settings & Configuration

Access settings via:

Ctrl+, (Windows/Linux) or Cmd+, (Mac) → Search "LocalAI"
Command Palette → "Preferences: Open Settings"

🧠 Recommended AI Models

O10D: LocalAI works seamlessly with any Ollama-compatible model. Here are the best use cases for the built-in model options:

Model	Category	Best Use Case
`gemma3:4b`	Balanced (Default)	Great balance of speed and performance. Ideal for everyday chat and code snippets.
`deepseek-coder:6.7b-instruct`	Coding & Agent	Excellent for deep logic and large-scale codebase explanations.
`qwen2.5-coder:7b`	Coding & Agent	Code generation, complex refactoring
`gemma3:270m`	General Chat	Tiny, Extremely lightweight and fast, ideal for quick completions and low-end hardware.
`gemma2:9b-instruct-q8_0`	General Chat	Strong reasoning capabilities, good for complex logic and thorough explanations.
`codeqwen:7b`	Coding	Solid alternative for code generation, debugging, and general programming tasks.
`qwen2.5-vl:7b`	Multi-modal	Advanced model with vision capabilities for image-related and multi-modal logic tasks.
`llama3.1:8b`	General Chat	Fast and highly capable for general programming questions, debugging, and explaining concepts.
`codegemma:7b`	Coding	Specialized for coding tasks, strong completion and generation abilities.
`starcoder2:7b`	Coding	Very fast, lightweight model perfect for real-time autocomplete with minimal latency.
`nomic-embed-text`	Embeddings (Default)	Highly recommended for indexing your workspace and performing semantic code searches.

🎯 Usage Examples

📝 Code Explanation

Select any code in your editor
Right-click → "LocalAI: Explain Selection"
Or use Ctrl+Shift+E / Cmd+Shift+E
Get detailed explanations in the chat panel! 🎉

🔧 Code Refactoring

Select code you want to improve
Right-click → "LocalAI: Refactor Selection"
Or use Ctrl+Shift+R / Cmd+Shift+R
Get optimized code suggestions! ✨

💬 General Chat

Open the chat panel (🤖 icon in sidebar)
Type your question or request
Add context files if needed (📎 button)
Get AI-powered assistance! 🤖

🎊 Ready to Code Smarter?

O10D: LocalAI is your personal AI coding assistant that respects your privacy while supercharging your development workflow! 🚀

Start coding with confidence - your AI assistant is ready to help! 🤖✨

A Hobby Project by Sreerag 🎯

O10D: LocalAI | Privacy-first Offline AI Chat & Coding Assistant

O10D