Skip to main content
The Cognite Toolkit has two components: the CLI (command-line interface) and the modules (YAML configuration files stored in module folders). You must keep both components in sync to ensure compatibility.

Prerequisites

  • Python 3.10 or later installed on your machine.
  • A Python package manager: pip, uv, or poetry.
  • The Cognite Toolkit already installed. If you haven’t installed it yet, see the setup guide.
1

Check the current version

To see the currently installed version of the Cognite Toolkit CLI, run this command:
pip show cognite-toolkit
Verify the installed CLI version:
cdf --version
To check the version of the modules, open the cdf.toml file and look for the [modules] section:
cdf.toml
...

[modules]
# This is the version of the modules. It should not be changed manually.
# It will be updated by the 'cdf modules upgrade' command.
version = "0.7.28"
...
2

Upgrade the Cognite Toolkit CLI

To upgrade the Cognite Toolkit to the latest version:
pip install --upgrade cognite-toolkit
To upgrade to a specific version:
pip install cognite-toolkit==0.7.59
Replace 0.7.59 with the version you want to install.
3

Upgrade the Cognite Toolkit modules

To upgrade the modules to match your installed CLI version:
  1. Run this command:
    cdf modules upgrade
    
  2. Follow the on-screen instructions to manually update any settings that couldn’t be upgraded automatically.
Run cdf --version and verify the output matches the version you installed. Then open cdf.toml and confirm the [modules] version has been updated.

Troubleshooting

”cdf command not found” error

A “command not found” error typically means your virtual environment is not activated. This is common when you use uv or poetry to manage Python dependencies in Visual Studio Code. If you haven’t installed the Cognite Toolkit yet, see the installation guide. Prefix your commands with uv run or poetry run to run them in the correct virtual environment:
uv run cdf modules upgrade
These tools don’t use pip run because pip doesn’t manage virtual environments. With pip, activate the virtual environment manually instead.
Or activate the virtual environment manually:
source path/to/venv/bin/activate  # Replace with your venv path

Mismatched installed and available versions

If you install a newer version of the Cognite Toolkit CLI using poetry and then run cdf --version in the terminal, the terminal may show an older version.
>> poetry add cognite_toolkit@latest
Using version ^0.7.59 for cognite-toolkit
>> cdf --version
CDF-Toolkit version: 0.6.40.
This happens because the terminal is not running in the same virtual environment as poetry. To fix this, either prefix your commands with poetry run or activate the virtual environment with poetry shell.

Next steps

Last modified on February 13, 2026