Windows Install

This guide walks through a complete installation of GCTRL on Windows. GCTRL runs as Docker containers; on Windows those containers run through Docker Desktop with the WSL 2 backend. You run the installer from inside a WSL 2 (Ubuntu) terminal and access the dashboard from your normal Windows browser.

For full performance you also install native Ollama for Windows (GPU-accelerated) and point GCTRL at it - see Infrastructure for the why behind this. The general install flow mirrors Installation, adapted for Windows.

Overview

Requirements

Step 1 - Install Docker Desktop

1. Download Docker Desktop for Windows from https://www.docker.com/products/docker-desktop and install it.
2. Launch Docker Desktop.
3. Open Settings → General and confirm Use the WSL 2 based engine is enabled.
4. Keep Docker Desktop running. The GCTRL containers run through it.

Docker Desktop must be running whenever you use GCTRL. If it is stopped, the platform will not start.

Step 2 - Install WSL 2 and Ubuntu

1. Open PowerShell as Administrator.

2. Install WSL with the default Ubuntu distribution:

   wsl --install
   

3. Reboot when prompted.

4. After reboot, finish the Ubuntu first-run setup (create your Linux username and password).

5. Back in Docker Desktop, open Settings → Resources → WSL Integration and enable integration for your Ubuntu distro. Apply and restart Docker Desktop if asked.

This lets Docker Desktop run containers inside the WSL 2 environment that your Ubuntu shell talks to.

Step 3 - Install native Ollama for Windows

This step unlocks GPU acceleration. The Ollama bundled inside Docker is CPU-only - Docker cannot reach your NVIDIA GPU. Installing Ollama natively on Windows lets GCTRL use the GPU for both the RAG/agent path and KEX extraction and embeddings. See Infrastructure for more detail.

1. Download Ollama for Windows (native, NVIDIA-GPU accelerated) from https://ollama.com/download.

2. Install and run it.

3. Open PowerShell and pull a model:

   ollama pull qwen2.5:7b
   

   On machines with more RAM/VRAM you can pull a larger model such as qwen2.5:14b or qwen2.5:32b.

You connect GCTRL to this native Ollama in step 6.

Step 4 - Run the GCTRL installer

1. Open an Ubuntu (WSL) terminal (search "Ubuntu" in the Start menu, or run wsl in PowerShell).

2. Run the installer:

   curl -fsSL https://gctrl.tech/install | bash
   

   Docker Desktop runs the containers via WSL 2, so no extra Docker setup is needed inside Ubuntu.

3. The installer is interactive on first run: it detects existing services, lets you pick a model, brings up the stack, and deploys the FUSE resolution engine automatically.

To skip the model picker for a scripted install, set the model up front:

GCTRL_MODEL=qwen2.5:7b curl -fsSL https://gctrl.tech/install | bash

Ports

Step 5 - Create your account and activate

1. Open http://localhost:3001 in your Windows browser.

2. Create your admin account.

3. Enter your license key in the format:

   GCTRL-XXXX-XXXX-XXXX-XXXX-XXXX
   

For more on licensing, see Activation.

Step 6 - Point GCTRL at native Ollama

This repoints both the RAG/agent path and KEX extraction and embeddings to your native, GPU-accelerated Ollama.

1. In the dashboard, go to Settings → Infrastructure.

2. Set the Ollama base URL to:

   http://host.docker.internal:11434
   

3. Save.

host.docker.internal is how containers reach a service running on your Windows host, so this points the in-container GCTRL at the native Windows Ollama from step 3.

Without this switch, GCTRL keeps using the bundled CPU-only Ollama - fine to start, slow for real work. See Infrastructure for details on the inference path.

Troubleshooting

Next steps

• Activation - create your admin account and enter your license key.
• Infrastructure - native Ollama, GPU, and swappable backing stores.
• Quickstart - your first extraction and graph.