Docs/Ollama Setup

Ollama Setup

NeuroTerm uses Ollama to run AI locally on your machine. No API keys, no cloud, complete privacy.

What is Ollama?

Ollama is an open-source tool that runs large language models (LLMs) locally on your computer. NeuroTerm connects to Ollama to power:

  • Magic Input — Natural language to terminal commands
  • Local RAG — Ask questions about your datasheets
  • AI Transforms — Intelligent output filtering

1. Install Ollama

Windows

Download and run the installer from ollama.com

Download Ollama for Windows

After installation, Ollama runs as a background service. You'll see the llama icon in your system tray.

2. Pull a Model

Open a terminal (cmd, PowerShell, or Windows Terminal) and pull a model:

Terminal
$ ollama pull llama3.2

Recommended Models

  • llama3.2 — Best balance of speed and quality (2GB)
  • qwen2.5:3b — Fast, good for Magic Input (2GB)
  • mistral — Great instruction following (4GB)

The download may take a few minutes depending on your connection.

3. Verify Installation

Test that Ollama is running:

Terminal
$ ollama list
NAME ID SIZE MODIFIED
llama3.2 a80c4f17acd5 2.0 GB 2 minutes ago

You should see your downloaded model in the list.

4. Configure NeuroTerm

Open NeuroTerm and go to Settings (gear icon):

Ollama Endpoint
http://localhost:11434
Default
Model
llama3.2
Your pulled model

Click "Test Connection" to verify NeuroTerm can reach Ollama.

Troubleshooting

"Connection refused" error

Make sure Ollama is running. Check your system tray for the llama icon, or run ollama serve manually.

Slow responses

Try a smaller model like qwen2.5:1.5b. If you have an NVIDIA GPU, Ollama will use it automatically for faster inference.

Model not found

Ensure the model name in NeuroTerm settings matches exactly what you pulled. Run ollama list to see available models.

Next Steps