Install or update the Databricks CLI

Note

This information applies to Databricks CLI versions 0.205 and above, which are in Public Preview. To find your version of the Databricks CLI, run databricks -v.

This article describes how to install or update the Databricks CLI. See What is the Databricks CLI?.

Install

To install Databricks CLI version 0.205 or above:

The Databricks CLI is also available from within the Databricks workspace user interface. See Run the Databricks CLI from within a Databricks workspace.

Update

To update an earlier installation of Databricks CLI version 0.205 or above:

Homebrew installation for Linux or macOS

For this installation option, you use Homebrew to automatically download and install the latest Databricks CLI executable release.

From your terminal, just run the following two Homebrew commands to install the CLI:

brew tap databricks/tap
brew install databricks

Then continue with Authentication for the Databricks CLI.

If you are not sure whether you have Homebrew installed, or if the preceding Homebrew commands fail, follow these longer instructions:

  1. Check whether Homebrew is already installed by running the following command from a Linux shell prompt or macOS Terminal. If Homebrew is installed, the Homebrew version number is displayed:

    brew -v

  2. If Homebrew is not already installed, install it by running the following command:

    /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

    Then run the following command to verify the Homebrew installation, which displays the Homebrew version number:

    brew -v

  3. Use Homebrew to add the databricks/homebrew-tap repository in GitHub to your list of available Homebrew Tap repositories, by running the following command:

    brew tap databricks/tap

  4. Use Homebrew to instruct the databricks/homebrew-tap repository to download and install the Databricks CLI executable, by running the following command:

    brew install databricks

  5. Confirm whether the Databricks CLI is installed correctly. To do this, view the Databricks CLI executable’s version by using the -v option or by running the version command:

    databricks -v

# Or:
databricks version

    If a version number of 0.205.0 or above is listed, it means that the Databricks CLI is installed correctly.

    Note

    If you run databricks but get an error such as command not found: databricks, or if you run databricks -v and a version number of 0.18 or below is listed, this means that your machine cannot find the correct version of the Databricks CLI executable. To fix this, see Verify your CLI installation.

  6. Continue with Authentication for the Databricks CLI.

WinGet installation for Windows

For this installation option, you use winget to automatically download and install the latest Databricks CLI executable release.

  1. From your Command Prompt, run the following two winget commands to install the CLI, and then restart your Command Prompt:

    winget search databricks
winget install Databricks.DatabricksCLI

  2. Confirm whether the Databricks CLI is installed correctly. To do this, view the Databricks CLI executable’s version by using the -v option or by running the version command:

    databricks -v

# Or:
databricks version

    If a version number of 0.205.0 or above is listed, it means that the Databricks CLI is installed correctly.

    Note

    If you run databricks but get an error such as command not found: databricks, or if you run databricks -v and a version number of 0.18 or below is listed, this means that your machine cannot find the correct version of the Databricks CLI executable. To fix this, see Verify your CLI installation.

  3. Continue with Authentication for the Databricks CLI.

Chocolatey installation for Windows

Note

Installing the Databricks CLI with Chocolatey is Experimental.

For this installation option, you use Chocolatey to automatically download and install the latest Databricks CLI executable release.

  1. From your Command Prompt, run the following choco command to install the CLI:

    choco install databricks-cli

  2. Confirm whether the Databricks CLI is installed correctly. To do this, view the Databricks CLI executable’s version by using the -v option or by running the version command:

    databricks -v

# Or:
databricks version

    If a version number of 0.205.0 or above is listed, it means that the Databricks CLI is installed correctly.

    Note

    If you run databricks but get an error such as command not found: databricks, or if you run databricks -v and a version number of 0.18 or below is listed, this means that your machine cannot find the correct version of the Databricks CLI executable. To fix this, see Verify your CLI installation.

  3. Continue with Authentication for the Databricks CLI.

curl installation for Linux, macOS, and Windows

For this installation option, you use curl to download and install the latest Databricks CLI executable release.

For Windows, Databricks recommends that you use Windows Subsystem for Linux (WSL). You must also have unzip installed. For installation instructions, see your operating system’s documentation. If you cannot use WSL, skip this procedure and use the Command Prompt or PowerShell to install the Databricks CLI from source instead.

  1. curl must be installed. You can check whether curl is already installed by running the following command from a Linux shell prompt, macOS Terminal, or Bash on WSL. If curl is installed, the curl version number is displayed:

    curl -V

  2. If curl is not already installed, install it by following the instructions on the curl Releases and Downloads page for your operating system and architecture. To get your machine’s architecture, see your operating system’s documentation.

    Then run the following command to verify the curl installation, which displays the curl version number:

    curl -V

  3. Use curl to download and install the Databricks CLI executable by running the following command (note that you must specify the correct uppercase and lowercase characters for -fsSL on all operating systems):

    curl -fsSL https://raw.githubusercontent.com/databricks/setup-cli/main/install.sh | sh

    Note

    To view the script’s contents before you run it, see the install.sh file in the Databricks CLI Setup repository in GitHub.

    For Windows, you should run this command as an administrator.

    For Linux and macOS, if an error message states that /usr/local/bin is not writable, then run the command again with sudo.

    This command downloads and installs the Databricks CLI executable in the path /usr/local/bin/databricks on macOS and Linux, and C:\Windows\databricks.exe on Windows.

    If for some reason the Databricks CLI is already installed, the following error appears: “Target path <path> already exists.” To fix this, you must manually delete the Databricks CLI executable from the preceding path, and then run the curl command again.

  4. Confirm whether the Databricks CLI is installed correctly. To do this, view the Databricks CLI executable’s version by using the -v option or by running the version command:

    databricks -v

# Or:
databricks version

    If a version number of 0.205.0 or above is listed, it means that the Databricks CLI is installed correctly.

    Note

    If you run databricks but get an error such as command not found: databricks, or if you run databricks -v and a version number of 0.18 or below is listed, this means that your machine cannot find the correct version of the Databricks CLI executable. To fix this, see Verify your CLI installation.

  5. Continue with Authentication for the Databricks CLI.

Source installation for Linux, macOS, and Windows

For this installation option, you manually download a .zip file and then manually extract the Databricks CLI executable from the downloaded .zip file.

  1. Download onto your local development machine the latest Databricks CLI .zip file, as listed in the Releases section of the Databricks CLI repository in GitHub. The latest release has a Latest label next to it in the list of releases. The latest pre-release has a Pre-release label and is more suitable for experimentation and early adopters. The .zip file that you download from the Assets section for the release must match your development machine’s operating system and architecture:

    Filename

    Architecture

    databricks_cli_X.Y.Z_darwin_amd64.zip

    macOS, Intel 64-bit / x86_64

    databricks_cli_X.Y.Z_darwin_arm64.zip

    macOS, Apple Silicon / M Series / arm64

    databricks_cli_X.Y.Z_linux_amd64.zip

    Linux, Intel 64-bit / x86_64

    databricks_cli_X.Y.Z_linux_arm64.zip

    Linux, ARM 64-bit / arm64

    databricks_cli_X.Y.Z_windows_386.zip

    Windows, Intel 32-bit / x86

    databricks_cli_X.Y.Z_windows_amd64.zip

    Windows, Intel 64-bit / AMD64

    databricks_cli_X.Y.Z_windows_arm64.zip

    Windows, ARM 64-bit / ARM64

    To get your machine’s architecture, see your operating system’s documentation. For example:

    • Linux: The terminal commands uname -m, lscpu, or sudo lshw -c cpu.

    • macOS: Apple icon menu > About This Mac, or the Terminal command uname -m.

    • Windows: The Command Prompt or PowerShell command echo %PROCESSOR_ARCHITECTURE%.

    If you need to verify the integrity of one of these .zip files, Databricks provides a checksum file named databricks_cli_X.Y.Z_SHA256SUMS in the same list as the .zip files. To run a checksum verification, see your operating system’s documentation.

  2. Extract the contents of the downloaded .zip file. To extract the .zip file, see your operating system’s documentation.

  3. In the extracted content, a folder appears with the same name as the .zip file. Inside of this folder is the Databricks CLI executable. You can leave the Databricks CLI executable there, or you can copy or move it to another location.

  4. Confirm whether the Databricks CLI is installed correctly. To do this, from a Linux shell prompt, macOS Terminal, or Windows Command Prompt, view the Databricks CLI executable’s version by using the -v option or by running the version command:

    databricks -v

# Or:
databricks version

    If a version number of 0.205.0 or above is listed, it means that the Databricks CLI is installed correctly.

    Note

    If you run databricks but get an error such as command not found: databricks, or if you run databricks -v and a version number of 0.18 or below is listed, this means that your machine cannot find the correct version of the Databricks CLI executable. To fix this, see Verify your CLI installation.

    For macOS, you might need to authorize your system to be able to run the Databricks CLI first. See Open a Mac app from an unidentified developer.

  5. Continue with Authentication for the Databricks CLI.

Homebrew update for Linux or macOS

For this update option, you use Homebrew to automatically update a previous installation of Databricks CLI version 0.205 or above to the latest version.

  1. From a Linux shell prompt or macOS Terminal, use Homebrew to download and update to the latest version of the Databricks CLI executable by running the following command:

    brew upgrade databricks

  2. Confirm whether the Databricks CLI is updated correctly. To do this, view the Databricks CLI executable’s version by using the -v option or by running the version command:

    databricks -v

# Or:
databricks version

    If the expected updated version number displays, then the Databricks CLI is installed correctly.

    For a list of Databricks CLI versions, see the Releases section of the Databricks CLI repository in GitHub. The latest release has a Latest label next to it in the list of releases.

    Note

    If you run databricks but get an error such as command not found: databricks, or if you run databricks -v and a version number of 0.18 or below is listed, this means that your machine cannot find the correct version of the Databricks CLI executable. To fix this, see Verify your CLI installation.

WinGet update for Windows

For this update option, you use winget to automatically download and update a previous version of Databricks CLI version 0.205 or above to the latest version.

  1. From your Command Prompt, use winget to download and update to the latest version of the Databricks CLI executable by running the following command:

    winget upgrade Databricks.DatabricksCLI

  2. Confirm whether the Databricks CLI is installed correctly. To do this, view the Databricks CLI executable’s version by using the -v option or by running the version command:

    databricks -v

# Or:
databricks version

    If the expected updated version number displays, then the Databricks CLI is installed correctly.

    For a list of Databricks CLI versions, see the Releases section of the Databricks CLI repository in GitHub. The latest release has a Latest label next to it in the list of releases.

    Note

    If you run databricks but get an error such as command not found: databricks, or if you run databricks -v and a version number of 0.18 or below is listed, this means that your machine cannot find the correct version of the Databricks CLI executable. To fix this, see Verify your CLI installation.

Chocolatey update for Windows

Note

Updating the Databricks CLI with Chocolatey is Experimental.

For this update option, you use Chocolatey to automatically download and update a previous version of Databricks CLI version 0.205 or above to the latest version.

  1. From your Command Prompt, use choco to download and update to the latest version of the Databricks CLI executable by running the following command:

    choco upgrade databricks-cli

  2. Confirm whether the Databricks CLI is installed correctly. To do this, view the Databricks CLI executable’s version by using the -v option or by running the version command:

    databricks -v

# Or:
databricks version

    If the expected updated version number displays, then the Databricks CLI is installed correctly.

    For a list of Databricks CLI versions, see the Releases section of the Databricks CLI repository in GitHub. The latest release has a Latest label next to it in the list of releases.

    Note

    If you run databricks but get an error such as command not found: databricks, or if you run databricks -v and a version number of 0.18 or below is listed, this means that your machine cannot find the correct version of the Databricks CLI executable. To fix this, see Verify your CLI installation.

curl update for Linux, macOS, and Windows

For this update option, you use curl to automatically download and update a previous installation of Databricks CLI version 0.205 or above to the latest version.

For Windows, Databricks recommends that you use Windows Subsystem for Linux (WSL). You must have unzip installed. For installation instructions, see your operating system’s documentation. If you cannot use WSL, you should skip this procedure and use the Command Prompt or PowerShell to update the Databricks CLI from source instead.

  1. Delete the existing installation of the Databricks CLI executable from the path /usr/local/bin/databricks on macOS and Linux, or C:\Windows\databricks.exe on Windows.

  2. From a Linux shell prompt, macOS Terminal, or Windows Command Prompt, use curl to download and install the latest version of the Databricks CLI executable by running the following command (note that you must specify the correct uppercase and lowercase characters for -fsSL):

    curl -fsSL https://raw.githubusercontent.com/databricks/setup-cli/main/install.sh | sh

    Note

    To view the script’s contents before you run it, see the install.sh file in the Databricks CLI Setup repository in GitHub.

    For Windows, you should run this command as an administrator.

    For Linux and macOS, if an error message states that /usr/local/bin is not writable, then run the command again with sudo.

    This command downloads and installs the Databricks CLI executable in the path /usr/local/bin/databricks on macOS and Linux, and C:\Windows\databricks.exe on Windows.

    If the following error appears, you must manually delete the Databricks CLI executable from the preceding path, and then run the curl command again: “Target path <path> already exists.”

  3. Confirm whether the Databricks CLI is updated correctly. To do this, from your terminal, view the Databricks CLI executable’s version by using the -v option or by running the version command:

    databricks -v

# Or:
databricks version

    If the expected updated version number displays, then the Databricks CLI is installed correctly.

    For a list of Databricks CLI versions, see the Releases section of the Databricks CLI repository in GitHub. The latest release has a Latest label next to it in the list of releases.

    Note

    If you run databricks but get an error such as command not found: databricks, or if you run databricks -v and a version number of 0.18 or below is listed, this means that your machine cannot find the correct version of the Databricks CLI executable. To fix this, see Verify your CLI installation.

Source update for Linux, macOS, and Windows

For this update option, you manually download a .zip file and then manually extract the Databricks CLI executable from the downloaded .zip file.

  1. Optionally, delete the Databricks CLI executable, the .zip file, and the .zip file’s extracted folder, from any previous manual source install procedure.

  2. Continue by following the instructions in Source installation for Linux, macOS, and Windows.

Run the Databricks CLI from within a Databricks workspace

You can access the Databricks CLI from the web terminal inside of a notebook that is connected to a running cluster, or directly from the web terminal for a running cluster. The available CLI is always the latest version, and authentication is based on the current user. The cluster must meet the following requirements:

  • The cluster must have Databricks Runtime 15.0 or above installed.

  • The cluster must use a cluster access mode of Single user or No isolation shared. Shared cluster access mode is not supported.

  • Serverless compute is not supported.

  • The workspace must not be enabled for Private Service Connect.

To access the CLI from the web terminal for a notebook, do the following:

  1. Connect the notebook to a cluster that meets the preceding requirements.

  2. At the very bottom of the notebook’s right sidebar, click the Open bottom panel icon Open bottom panel in notebook.

  3. Use the CLI in the Terminal pane that appears. For example, the following command prints information about the current user:

    databricks current-user me

To access the CLI from the web terminal for a cluster, do the following:

  1. On the workspace’s sidebar, click Compute.

  2. On the All-purpose compute tab, click the name of a cluster that meets the preceding requirements.

  3. Click Start.

  4. After the cluster starts, on the Apps tab, click Web Terminal.

  5. Use the CLI in the Databricks web terminal that appears. For example, the following command prints information about the current user:

    databricks current-user me