OpenClaude VS Code

AI coding assistant for VS Code powered by any LLM — OpenAI GPT-4o, Google Gemini, DeepSeek, Ollama, Codex, AWS Bedrock, and 200+ models via OpenAI-compatible APIs.

OpenClaude wraps the OpenClaude CLI in a full-featured VS Code extension with a chat interface, diff viewer, @-mentions, slash commands, session management, and more.

Features

Chat Interface

• Streaming chat panel with markdown rendering and syntax-highlighted code blocks
• Tool call visualization (collapsible blocks showing what the AI is doing)
• Session history — browse, resume, and fork past conversations
• Stop/interrupt generation at any time

Code Editing

• Native VS Code diff viewer for AI-proposed changes (accept/reject)
• @-mention files, folders, and line ranges for context
• Checkpoint/rewind — snapshot and restore to any point in conversation

Multi-Provider Support

Switch between LLM providers with /provider or the provider badge:

Developer Tools

• 5 permission modes (Default, Plan, Accept Edits, Bypass, Don't Ask)
• MCP (Model Context Protocol) server integration
• Plugin manager for MCP plugins
• Git worktree support for parallel AI sessions
• Status bar with live token count and cost
• Onboarding walkthrough for new users

Installation

From the VS Code Marketplace

Search for OpenClaude in the Extensions panel (Ctrl+Shift+X) and click Install.

From a .vsix file

code --install-extension openclaude-vscode-0.1.0.vsix

Prerequisites

Install the OpenClaude CLI first:

npm install -g @gitlawb/openclaude

Quick Start

1. Install the CLI

npm install -g @gitlawb/openclaude

2. Configure a provider

OpenAI (recommended):

export CLAUDE_CODE_USE_OPENAI=1
export OPENAI_API_KEY=sk-your-key-here
export OPENAI_MODEL=gpt-4o

Ollama (local, free):

ollama serve  # start Ollama first
export CLAUDE_CODE_USE_OPENAI=1
export OPENAI_API_KEY=ollama
export OPENAI_BASE_URL=http://localhost:11434/v1
export OPENAI_MODEL=llama3

Gemini:

export CLAUDE_CODE_USE_GEMINI=1
export GOOGLE_API_KEY=AIza-your-key
export GEMINI_MODEL=gemini-2.0-flash

Codex (ChatGPT):

export CLAUDE_CODE_USE_OPENAI=1
export OPENAI_API_KEY=sk-your-key
export OPENAI_BASE_URL=https://api.codex.openai.com/v1
export OPENAI_MODEL=gpt-5.4

Or use the /provider command in the chat to set up providers interactively.

3. Open OpenClaude

• Press Cmd+Escape (Mac) / Ctrl+Escape (Windows/Linux)
• Or click the OpenClaude icon in the sidebar
• Or run OpenClaude: Open in New Tab from the Command Palette

4. Start coding

Type a message and press Enter. Use @ to mention files, / for slash commands.

Keyboard Shortcuts

Slash Commands

Type / in the chat input to see all available commands. Key commands include:

Settings

Contributing

git clone https://github.com/Harsh1210/openclaude-vscode
cd openclaude-vscode
npm install
cd webview && npm install && cd ..
npm run build

Development (watch mode):

npm run watch
# Press F5 in VS Code to launch Extension Development Host

Package:

npx @vscode/vsce package --no-dependencies --allow-missing-repository

Architecture

Webview (React + Tailwind)
  |  postMessage
Extension Host (TypeScript)
  |  stdin/stdout NDJSON
OpenClaude CLI (child process)
  |  OpenAI Chat Completions API
Any LLM Provider

The extension is a thin UI wrapper. All intelligence (tools, providers, slash commands, MCP, plugins) lives in the CLI.

License

MIT