Install

Install Ubik

Free for personal, non-commercial and evaluation use. Closed source. For commercial use, see licensing and pricing.

Linux x86_64 and arm64, macOS arm64, macOS 12 or later. Install with Homebrew, pip or a tarball.

Try the live shell without installing →

homebrew
brew tap magiclex/ubikbrew trust magiclex/ubikbrew install magiclex/ubik/ubikubik versionubik 1.3.0ubik-engine 1.3.0

Homebrew requires the trust step for any third-party tap. Update with brew upgrade ubik.

Run a windowed count against our public demo stream. Each row is a merchant’s order count for a completed five-second window. Press Ctrl+C to stop.

Count orders on the demo stream
ubik --from kafka://demo.getubik.dev/orders \  "SELECT merchant, TUMBLE(ts, INTERVAL '5' SECOND) AS w,          count(*) AS orders   FROM orders GROUP BY merchant, TUMBLE(ts, INTERVAL '5' SECOND)"

For Python, install ubik-sql, import ubik.

python
pip install ubik-sql

Python quickstart →

Tarballs: Linux and macOS

Download and verify the tarball for your platform.

linux x86_64
curl -fsSLO https://getubik.dev/releases/v1.3.0/ubik-1.3.0-linux-amd64.tar.gzcurl -fsSL https://getubik.dev/releases/v1.3.0/SHA256SUMS \    | sha256sum -c --ignore-missingubik-1.3.0-linux-amd64.tar.gz: OKtar -xzf ubik-1.3.0-linux-amd64.tar.gz./bin/ubik versionubik 1.3.0ubik-engine 1.3.0
linux arm64
curl -fsSLO https://getubik.dev/releases/v1.3.0/ubik-1.3.0-linux-arm64.tar.gzcurl -fsSL https://getubik.dev/releases/v1.3.0/SHA256SUMS \    | sha256sum -c --ignore-missingubik-1.3.0-linux-arm64.tar.gz: OKtar -xzf ubik-1.3.0-linux-arm64.tar.gz./bin/ubik versionubik 1.3.0ubik-engine 1.3.0
macos arm64
curl -fsSLO https://getubik.dev/releases/v1.3.0/ubik-1.3.0-darwin-arm64.tar.gzcurl -fsSL https://getubik.dev/releases/v1.3.0/SHA256SUMS \    | shasum -a 256 -c --ignore-missingubik-1.3.0-darwin-arm64.tar.gz: OKtar -xzf ubik-1.3.0-darwin-arm64.tar.gz./bin/ubik versionubik 1.3.0ubik-engine 1.3.0
Intel Mac support

Intel macs are not built. Leave an email and you get one message if that changes, nothing else.

Next: create a checkpointed pipeline on your own broker. Configure the source, output and SQL, then start it with the CLI.