Skip to main content

Windows Installation

Quick SummaryInstalling Quanty AI on Windows

Quanty AI is build for Windows 10 and 11 for now. By combining the Ollama backend with NVIDIA CUDA acceleration, you will get a great local AI companion experience that stays entirely on your hardware. Linux and MacOS will be supported in future releases.

Source: Quanty AI Knowledge Base
Quanty
Quanty

Welcome to Windows! This is where I'm most at home right now. Follow these steps to get me running with full GPU power so I can think as fast as possible!

System Requirements​

To ensure a smooth experience with local models, your system should meet these minimum specifications.

Compatibility Checklist​

  • Operating System: Windows 10 22H2 or newer, Home or Pro
  • NVIDIA Drivers: NVIDIA 531 or newer Drivers if you have an NVIDIA card.
  • AMD Radeon Drivers: AMD Radeon Driver https://www.amd.com/en/support if you have a Radeon card
  • Hardware Support: Since Hardware support recommendation is based on Ollama, visit https://docs.ollama.com/gpu for further information.

Step 1: Install Dependencies​

Quanty AI relies on Ollama to serve local models and NVIDIA CUDA for acceleration.

1.1 Install Ollama​

Ollama is the engine that powers the text-based reasoning (like Qwen 3 or Mistral).

  1. Go to Ollama.com.
  2. Download the Windows installer.
  3. Run the .exe and follow the prompts.

1.2 Update NVIDIA Drivers​

To use your GPU for AI, you need modern drivers.

  1. Go to the NVIDIA Driver Downloads.
  2. Install version 531 or higher.
  3. Verify by opening your terminal and typing: 
    nvidia-smi

1.3 Change Model Storage (Optional)​

If you have limited space on your C: drive, you can tell Ollama to store my "brain" files elsewhere.

  1. Search for "Environment Variables" in the Windows Start menu.
  2. Click Edit environment variables for your account.
  3. Click New... under User variables.
  4. Variable name: OLLAMA_MODELS
  5. Variable value: D:\AI\Models (or your preferred path).
  6. Restart your computer to apply.

Step 2: Install Quanty AI​

  1. Download: Use the link from your Gumroad confirmation email.
  2. Run Installer: Launch the QuantyAI-Setup.exe.
  3. Permissions: Windows may show a "SmartScreen" warning since we are a new publisher. Click More Info > Run Anyway.
🖼️
Quanty AI Installation Wizard
[Visual Coming Soon]

Step 3: Verify GPU Acceleration​

Once Quanty is open, let's make sure I'm using your graphics card instead of your CPU.

  1. Open Settings (Gear icon in the sidebar).
  2. Scroll to the bottom of the Settings page.
  3. Check if your NVIDIA GPU is listed. You will also see how much VRAM you are using.
Quanty
Quanty

If you see your GPU's name in the settings, I'm ready to use my full "brain power"! If not, double-check that your NVIDIA drivers are up to date. Oh, and if you have a Radeon GPU card, it might work even if you don't see it in the settings page. I was focused on NVIDIA cards system information.

Troubleshooting & Logs​

If something isn't working right, checking the logs is the best way to see what's happening under the hood.

  • Access Logs: Paste %LOCALAPPDATA%\Ollama into your File Explorer address bar.
  • Important Files:
    • app.log: Logs from the Ollama application.
    • server.log: Logs from the model engine (very useful for GPU errors).
  • Slow Performance: Check if other apps are using your VRAM (like Chrome or Games).
  • AVX2 Errors: If your CPU is older than 2013, models may run very slowly or fail to load.
Quick SummaryReady to Begin?

Now that you're installed, head over to the Quickstart to pull your first model and start chatting!

Source: Quanty AI Knowledge Base