K3ai is a lightweight tool to get an AI Infrastructure Stack up in minutes not days.

NOTE on the K3ai origins

Original K3ai Project has been developed at the end of October 2020 in 2 weeks by:

K3ai v1.0 has been entirely re-written by Alessandro Festa during the month of October 2021 to
offer a better User Experience.

Thanks to the amazing and incredible people and projects that have been instrumental to create K3ai project repositories,website,etc…

⚡️ Quick start

Let’s discover K3ai in three simple steps.

🌘 Getting Started

Get started by download k3ai from the release page here.

Or try K3ai companion script using this command:

curl -LO https://get.k3ai.in | sh -

🌗 Load K3ai configuration

Let’s start loading the configuration:

k3ai up

First time k3ai run will ask for a Github PAT (Personal Access Token) that we will use to avoid API calls limitations. Check Github Documentation to learn how to create one. Your personal GH PAT only need read repository permission.

🌖 Configure the base infrastructure

Choose your favourite Kubernetes flavor and run it:

To know which K8s flavors are available

k3ai cluster list --all

it should print something like:

│ INFRASTRUCTURE                                                                                          │
│ TYPE  │ DESCRIPTION                                         │ KIND  │ TAG    │ VERSION │ STATUS         │
│ CIVO  │ The First Cloud Native Service Provider Power...    │ infra │ cloud  │ latest  │ Available      │
│ EKS-A │ Amazon Eks Anywhere Is A New Deployment Option...   │ infra │ hybrid │ v0.5.0  │ Available      │
│       │ ate And Operate Kubernetes Clusters On Custome...   │       │        │         │                │
│ K3S   │ K3s Is A Highly Available, Certified Kubernetes...  │ infra │ local  │ latest  │ Available      │
│       │ oads In Unattended, Resource-Constrained...         │       │        │         │                │
│ KIND  │ Kind Is A Tool For Running Local Kubernetes...      │ infra │ local  │ v0.11.2 │ Available      │
│       │ as Primarily Designed For Testing Kubernetes...     │       │        │         │                │
│       │  Or Ci.                                             │       │        │         │                │
│ TANZU │ Tanzu Community Edition Is A Fully-Featured...      │ infra │ hybrid │ latest  │ In Development │
│       │ ers And Users. It Is A Freely Available...          │       │        │         │                │
│       │  Of Vmware Tanzu.                                   │       │        │         │                │

Now let start with something super fast and super simple:

k3ai cluster deploy --type k3s --n mycluster

🌝 Install a plugin to do your AI experimentations

Now that the server is up and running let’s type:

k3ai plugin deploy -n mlflow -t mycluster

K3ai will print the url where you may access to the MLFLow tracking server at the end of the installation.
That’s all now just start having fun with K3ai!

🌈 Push a piece of code to the AI tools and focus on your goals

Let’s push some code to the AI tool (i.e.: MLFlow)

k3ai run --source https://github.com/k3ai/quickstart --target mycluster --backend mlflow

wait the run to complete and login the backend AI tolls (i.e.: on the MLFlow UI http://<YOUR IP>:30500)

Current Implementation support

Operating Systems

Operating System K3ai v1.0.0
Linux Yes
Windows In Progress
MacOs In Progress
Arm In Progress


K8s Clusters K3ai v1.0.0
Rancher K3s Yes
Vmware Tanzu Community Ed. Yes
Amazon EKS Anywhere Yes
KinD Yes


Plugins K3ai v1.0.0
Kuebflow Components Yes
MLFlow Yes
Apache Airflow Yes
Argo Workflows Yes

⭐️ Project assistance

If you want to say thank you or/and support active development of K3ai Project:

Together, we can make this project better every day! 😘

⚠️ License

K3ai is free and open-source software licensed under the BSD 3-Clause. Official logo was created by Alessandro Festa.


View Github