The Nuclio CLI (nuctl)

On This Page


nuctl is Nuclio's command-line interface (CLI), which provides command-line access to Nuclio features.

Download and Installation

To install nuctl, simply visit the Nuclio releases page and download the appropriate CLI binary for your environment (for example, nuctl-<version>-darwin-amd64 for a machine running macOS).

You can use the following command to download the latest nuctl release:

curl -s \
			| grep -i "browser_download_url.*nuctl.*$(uname)" \
			| cut -d : -f 2,3 \
			| tr -d \" \
			| wget -O nuctl -qi - && chmod +x nuctl


After you download nuctl, use the --help option to see full usage instructions:

nuctl --help

The running platform

nuctl automatically identifies the platform from which it's being run. You can also use the --platform CLI flag to ensure that you're running against a specific platform, as explained in this reference:

Local Docker

To force nuctl to run locally, using a Docker daemon, add --platform local to the CLI command.

For an example of function deployment using nuctl against Docker, see the Nuclio Docker getting-started guide.


To force nuctl to run against a Kubernetes instance of Nuclio, add --platform kube to the CLI command.

When running on Kubernetes, nuctl requires a running registry on your Kubernetes cluster and access to a kubeconfig file.

For an example of function deployment using nuctl against a Kubernetes cluster, see the Nuclio Kubernetes getting-started guide.

Invoking a function using nuctl on Kubernetes

While in most cases using nuctl on the local Docker platform enables you to invoke your function by running the nuctl invoke command on the host, on a Kubernetes platform this is a bit trickier and will most likely not work by default for some configurations. This means, for example, that the nuctl invoke command won't work unless your function is explicitly exposed or is reachable from wherever you run nuctl invoke. For more information, see Exposing a function.

For your convenience, when deploying a function using nuctl you can easily expose the function via a NodePort by using the --http-trigger-service-type=nodePort CLI option.