Skip to content

Installation

The easiest way to get started. Download, install, launch — no terminal needed.

  1. Download the latest .dmg from the Releases page.
  2. Open the .dmg and drag Knowmarks to your Applications folder.
  3. Double-click to launch.

macOS Gatekeeper

Knowmarks is not yet signed with an Apple Developer certificate. On first launch, macOS may block the app with "Knowmarks can't be opened because Apple cannot check it for malicious software."

To open it: right-click (or Control-click) the app in Applications, select Open, and click Open in the dialog. You only need to do this once — subsequent launches work normally.

The desktop app bundles everything: Python runtime, embedding model, and the full dashboard. A setup wizard walks you through API key configuration and importing your first bookmarks.

pip

Knowmarks requires Python 3.11+.

pip install 'knowmarks[all,embeddings]'

Includes local ONNX embeddings via fastembed (~500MB model download on first run). No external API needed.

pip install 'knowmarks[all]'

Use with Ollama or an OpenAI-compatible API for embeddings.

uv pip install 'knowmarks[all,embeddings]'

After installation, the km command is available globally.

Docker

Docker images are available in two variants:

docker run -v knowmarks_data:/data -p 3749:3749 knowmarks

docker run -v knowmarks_data:/data -p 3749:3749 \
  -e KNOWMARKS_EMBEDDING_PROVIDER=ollama \
  -e KNOWMARKS_EMBEDDING_ENDPOINT=http://host.docker.internal:11434 \
  knowmarks:slim

Or use Docker Compose:

docker compose up -d

The dashboard is available at http://localhost:3749.

Verify Installation

km status

This prints your collection statistics. If you see output, Knowmarks is installed correctly.

First run

On first run with the embeddings extra, Knowmarks downloads the BAAI/bge-small-en-v1.5 model (~130MB). This happens once and may take a minute on slow connections.