Note67 LogoNOTE67

OLLAMA SETUP GUIDE

Run AI models locally on your Mac. This guide will help you set up Ollama to power Note67's AI summarization features.

01

DOWNLOAD OLLAMA

Visit the official Ollama website to download the macOS app.

Download Ollama

After downloading:

  1. Open the downloaded .dmg file
  2. Drag Ollama to your Applications folder
  3. Launch Ollama from Applications
02

DOWNLOAD A MODEL (TERMINAL)

Open Terminal and run the following command to download your first model. We recommend starting with gemma3 for a good balance of speed and quality.

$ ollama pull gemma3

Other recommended models:

ollama pull gemma3:1b

Fastest, ~1GB • Good for quick summaries

ollama pull gemma3:4b

Balanced, ~3GB • Recommended for most users

ollama pull gemma3:12b

Higher quality, ~8GB • Better understanding

ollama pull llama3.2

Alternative model, ~2GB • Meta's latest

03

DOWNLOAD A MODEL (UI)

Prefer a graphical interface? Ollama has a built-in chat UI where you can browse and download models.

Ollama UI showing model selection with gemma3 variants

Ollama's built-in UI showing available gemma3 model variants

How to use the UI:

  1. Click the Ollama icon in your menu bar
  2. Select "Open Ollama" to launch the chat interface
  3. Click the model dropdown at the bottom
  4. Type to search for models (e.g., "gemma3")
  5. Click the download icon next to a model to install it
  6. Models with no icon are already installed
04

VERIFY INSTALLATION

Confirm Ollama is running and your models are installed.

$ ollama listNAME SIZE MODIFIED
gemma3:latest 3.0 GB 2 hours ago
gemma3:1b 815 MB 1 hour ago

You should see a list of your installed models. Ollama runs as a background service, so it will be available whenever Note67 needs it.

05

CONNECT TO NOTE67

Note67 automatically detects Ollama when it's running. Just make sure Ollama is active in your menu bar before using AI features.

In Note67:

  1. Open Note67 Settings
  2. Navigate to the AI section
  3. Select your preferred Ollama model
  4. Start transcribing and enjoy AI-powered summaries!

TIPS

MEMORY USAGE

Larger models produce better results but use more RAM. Start with smaller models (1b/4b) and upgrade if needed.

MULTIPLE MODELS

You can install multiple models and switch between them in Note67 depending on your needs.

UPDATES

Run ollama pull [modelname] again to update a model to its latest version.

PRIVACY

All processing happens locally on your Mac. Your transcripts and summaries never leave your device.

Back to home