⚡ Infinicode

AI coding agent powered by your local Ollama. Fork of OpenCode optimized for self-hosted LLMs.

npm install -g infinicode

# Connect to your Ollama master
infinicode connect 192.168.1.100

# Start coding
infinicode

Why Infinicode?

• 100% Local — All inference on YOUR hardware
• Zero Cloud — No API keys, no subscriptions, no data leaving your network
• One Command Setup — infinicode connect <ip> and you're done
• Multi-Node — Run on any machine, point to one Ollama master

Quick Start

1. Install

npm install -g infinicode
# or
pnpm add -g infinicode

2. Setup Ollama Master

On your GPU machine:

# Allow network connections
OLLAMA_HOST=0.0.0.0:11434 ollama serve

# Pull a coding model
ollama pull qwen2.5-coder:14b

3. Connect & Code

# From any machine on your network
infinicode connect 192.168.1.100   # Your Ollama IP
infinicode                          # Start coding!

Commands

| Command | Alias | Description | |---------|-------|-------------| | infinicode connect <ip> | ic c | Quick connect to Ollama master | | infinicode setup | ic s | Interactive setup wizard | | infinicode run | ic | Start the coding agent (default) | | infinicode models | ic m | List available models | | infinicode status | | Show config & connection status | | infinicode config --list | | View all configuration | | infinicode config --reset | | Reset to defaults |

Configuration

Config is stored automatically. Override with:

# Set default model
infinicode config --set defaultModel=codestral:22b

# Set master URL
infinicode config --set masterUrl=http://192.168.1.100:11434

# View all
infinicode config --list

Recommended Models

For coding tasks on Ollama:

| Model | Size | Best For | |-------|------|----------| | qwen2.5-coder:14b | 9GB | Great balance | | qwen2.5-coder:32b | 20GB | Best quality | | deepseek-coder-v2:16b | 10GB | Strong reasoning | | codestral:22b | 13GB | Fast completion |

Architecture

┌─────────────────┐
│   Laptop        │
│  (infinicode)   │──┐
└─────────────────┘  │
                     │    ┌─────────────────┐
┌─────────────────┐  │    │   GPU Server    │
│   Desktop       │──┼───▶│    (Ollama)     │
│  (infinicode)   │  │    │                 │
└─────────────────┘  │    │ qwen2.5-coder   │
                     │    │ codestral       │
┌─────────────────┐  │    └─────────────────┘
│   Server        │──┘
│  (infinicode)   │
└─────────────────┘

Security

Exposing Ollama to your network means anyone on that network can use it.

Secure options:

1. VPN/Tailscale — Only accessible on private network
2. SSH Tunnel — ssh -L 11434:localhost:11434 gpu-server
3. Firewall — Allow only specific IPs

Requirements

• Node.js 20+
• OpenCode installed (npm install -g opencode)
• Ollama running somewhere on your network

License

MIT — Based on OpenCode

⚡ Built for sovereign computing. Your code, your models, your hardware.