Dependencies Setup

Optional dependencies and tools to enhance your Clark experience.

System Requirements

PDF Processing with Poppler

Clark can process PDF files and convert them to images for visual understanding. This feature requires Poppler, a PDF rendering library that includes the pdftoppm utility.

Installing Poppler

Installation varies by operating system:

macOS (Homebrew)

$ brew install poppler

Ubuntu/Debian

$ sudo apt-get install poppler-utils

Fedora/RHEL

$ sudo dnf install poppler-utils

Windows

Download the latest release from the Poppler for Windows repository, extract it, and add the bin/ folder to your system PATH.

Verify Installation

After installation, verify that pdftoppm is available:

$ pdftoppm -v

You should see version information. If the command is not found, ensure Poppler's bin directory is in your PATH.

Note: If Poppler is not installed, Clark will still extract text from PDFs but won't be able to perform visual analysis of diagrams, equations, or images within the PDF.
Local AI Models

Running Models Locally with Ollama

Ollama lets you run large language models locally on your machine without requiring an API key or internet connection. This is perfect for:

  • Working offline or with intermittent connectivity
  • Keeping your data completely private
  • Avoiding API costs for frequent usage
  • Experimenting with different open-source models

Installing Ollama

  1. Visit ollama.ai and download the installer for your platform
  2. Run the installer and follow the setup instructions
  3. Open a terminal and verify the installation:
    4. $ ollama --version

Downloading Models

Once Ollama is installed, you need to download a model. Just make sure that the model supports tool use and image input.

Using Ollama with Clark

  1. Make sure Ollama is running (it starts automatically on most systems)
  2. Start Clark and use the /model command
  3. Select "Ollama" as your provider and choose your downloaded model
Tip: You can run multiple models and switch between them. Use /model anytime to change which model Clark uses.
Performance Note: Local models require significant RAM and CPU/GPU resources. For the best experience, ensure your machine has at least 8GB of RAM and consider models sized appropriately for your hardware.
Note Management

Working with Obsidian

Obsidian is a powerful markdown editor that works perfectly with Clark's note structure. Since Clark stores everything as plain markdown files, you can use Obsidian to:

  • View and edit your notes with a rich markdown preview
  • Visualize connections between notes with the graph view
  • Use plugins for advanced features like daily notes, templates, and spaced repetition
  • Search across all your notes with powerful query syntax
  • Work offline with full sync across devices

Setting Up Obsidian with Clark

  1. Download and install Obsidian from obsidian.md
  2. Open Obsidian and select "Open folder as vault"
  3. Navigate to your Clark workspace directory (e.g., ~/Library)
  4. Obsidian will now show all your markdown notes, wikilinks, and folder structure

Recommended Workflow

Many users find it helpful to use both Clark and Obsidian together:

  • Use Clark for interactive learning, asking questions, and working through problems
  • Use Obsidian for reviewing notes, organizing content, and seeing the big picture
  • Edit templates and structures in Obsidian's rich editor
  • Use Obsidian's graph view to discover connections you didn't realize existed
Tip: In Obsidian settings, enable "Automatically update internal links" and "Use [[Wikilinks]]" to ensure your links work seamlessly between Clark and Obsidian.
Note: Obsidian is completely optional. Clark works perfectly fine on its own — Obsidian just provides a different way to interact with the same markdown files.
Next Steps

You're All Set

These dependencies are all optional. Clark will work great with just the basic installation — these tools simply enhance specific features.

← Back to Getting Started