Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Important
Omnigent is in Beta. For the open-source documentation, see the Omnigent documentation.
Omnigent runs as a managed deployment on Azure Databricks: the server that coordinates your sessions is managed for you, and you sign in with your Azure Databricks workspace identity. This page covers the setup that is specific to Azure Databricks. See the open-source Omnigent documentation for more information about Omnigent features and functionality.
Prerequisites
You must have:
- The Omnigent preview enabled for your workspace. See Manage workspace-level previews.
- A workspace in a region that supports Databricks Unity AI Gateway. See Databricks feature availability by region.
To install Omnigent using the CLI, you must have:
- Python 3.12+
- Node.js 22 LTS with npm and tmux
Launch Omnigent on your own machine
Install Omnigent on your laptop or a Azure Databricks VM, and register it as a host.
Install Omnigent with the
databricksextra, which adds the managed-platform integration.Install script
curl -fsSL https://omnigent.ai/install.sh | sh -s -- --extra "databricks"UV
uv tool install "omnigent[databricks]"Pip
pip install "omnigent[databricks]"Set up model credentials. Run
omni setup, then add a credential for one or more agent harnesses:omni setup
- To route models through AI Gateway, choose Databricks in the credential picker and enter your workspace URL.
- Otherwise, provide your own API key, subscription, or gateway.
Sign in with your Azure Databricks identity and register the machine as an Omnigent host:
omni login <workspace-url> omni host --server <workspace-url>In your web browser, navigate to
<workspace-url>/omnigent.Select New session.
Open the host picker and select the machine you registered.
Describe a task in the composer and send it to start the session.
Connect the desktop app
The Omnigent desktop app is optional and is currently available on macOS only. Install the desktop app, then, once you have created a host, select Connect to new server in the app and enter your workspace URL.