OpenClaude 🤖

Run Claude Code with your own local AI models via NVIDIA NIM

OpenClaude is a local AI integration layer that lets you use Claude Code with NVIDIA NIM models (and other OpenAI-compatible APIs) instead of Anthropic's hosted API.

Why OpenClaude?

Feature	Claude Code (Default)	OpenClaude
LLM Provider	Anthropic only	Your choice
Cost	Per-token pricing	Your GPU / API
Privacy	Sends data to Anthropic	Local only
Models	Claude models	Any NIM model

Quick Start

# 1. Clone the repo
git clone https://github.com/Hanishchow/OpenClaude.git
cd OpenClaude

# 2. Configure your NVIDIA API key
cp config/env.example config/.env
# Edit .env and add: NVIDIA_NIM_API_KEY=your-key

# 3. Start the proxy
./scripts/start.sh

# 4. Use the CLI
./cli/nim-chat.sh " Say hello"

Architecture

┌─────────────────────────────────────────────────────────────┐
│                    Claude Code                              │
│                  (your IDE integration)                      │
└─────────────────────┬───────────────────────────────────────┘
                      │ UserPromptSubmit Hook
                      ▼
┌─────────────────────────────────────────────────────────────┐
│              OpenClaude Proxy (free-claude-code)            │
│               http://localhost:8082                        │
└─────────────────────┬───────────────────────────────────────┘
                      │
                      ▼
┌─────────────────────────────────────────────────────────────┐
│                  NVIDIA NIM                                 │
│         https://integrate.api.nvidia.com/v1              │
│                                                               │
│  Models: z-ai/glm-4.7, meta/llama-3.1-70b-instruct, etc.  │
└─────────────────────────────────────────────────────────────┘

Projects

Directory	Purpose
/cli	Command-line tools
/scripts	Startup and utility scripts
/config	Configuration templates
/docs	Documentation

CLI Usage

# Chat with NIM model
./cli/nim-chat.sh "Your prompt here"

# Route prompt to optimal model
./cli/nim-route.sh "description of task"

Environment Variables

Variable	Required	Description
NVIDIA_NIM_API_KEY	Yes	Your NVIDIA API key
NIM_BASE_URL	No	NIM endpoint (default: https://integrate.api.nvidia.com/v1)
FCC_PORT	No	Proxy port (default: 8082)

Get Your NVIDIA API Key

1. Go to https://build.nvidia.com/settings/api-keys
2. Create a new API key
3. Add to your .env file

Requirements

• Python 3.14+
• uv (package manager)
• NVIDIA API key

Contributing

See CONTRIBUTING.md for setup and development guidelines.

License

MIT License - See LICENSE

---

Note: This project is not affiliated with Anthropic or NVIDIA. It's a community integration for running Claude Code with alternative LLM providers.