Skip to content

What is dstack?

dstack is a streamlined alternative to Kubernetes, specifically designed for AI. It simplifies container orchestration for AI workloads both in the cloud and on-prem, speeding up the development, training, and deployment of AI models.

dstack is easy to use with any cloud providers as well as on-prem servers.

Accelerators

dstack supports NVIDIA GPU, AMD GPU, and Google Cloud TPU out of the box.

How does it work?

1. Set up the server

Before using dstack, ensure you've installed the server, or signed up for dstack Sky .

2. Define configurations

dstack supports the following configurations:

  • Dev environments — for interactive development using a desktop IDE
  • Tasks — for scheduling jobs (incl. distributed jobs) or running web apps
  • Services — for deployment of models and web apps (with auto-scaling and authorization)
  • Fleets — for managing cloud and on-prem clusters

Configuration can be defined as YAML files within your repo.

3. Apply configurations

Apply the configuration either via the dstack apply CLI command (or through a programmatic API.)

dstack automatically manages infrastructure provisioning and job scheduling, while also handling auto-scaling, port-forwarding, ingress, and more.

Why dstack?

dstack's founder and CEO explains the challenges dstack addresses for AI and Ops teams.

dstack streamlines infrastructure management and container usage, enabling AI teams to work with any frameworks across cloud platforms or on-premise servers.

How does it compare to other tools?

Kubernetes

How does dstack compare to Kubernetes?

dstack and Kubernetes are both container orchestrators for cloud and on-premises environments.

However, dstack is more lightweight, and is designed specifically for AI, enabling AI engineers to handle development, training, and deployment without needing extra tools or Ops support.

With dstack, you don't need Kubeflow or other ML platforms on top—everything is available out of the box.

Additionally, dstack is much easier to use for on-premises servers—just provide hostnames and SSH credentials, and dstack will automatically create a fleet ready for use with development environments, tasks, and services.

How does dstack compare to KubeFlow?

dstack can be used entirely instead of Kubeflow. It covers everything that Kubeflow does, and much more on top, including development environments, services, and additional features.

dstack is easier to set up with on-premises servers, doesn't require Kubernetes, and works with multiple cloud providers out of the box.

Can dstack and Kubernetes be used together?

For AI development, it’s more efficient to use dstack directly with your cloud accounts or on-prem servers—without Kubernetes.

However, if you prefer, you can set up the dstack server with a Kubernetes backend to provision through Kubernetes.

Does your Ops team insist on using Kubernetes for production-grade deployment? You can use dstack and Kubernetes side by side; dstack for development and Kubernetes for production-grade deployment.

Slurm

How does dstack compare to Slurm?

dstack can be used entirely instead of Slurm. It covers everything that Slurm does, and a lot more on top, including dev environments, services, out-of-the-box cloud support, easier setup with on-premises servers, and much more.

Where do I start?

  1. Proceed to installation
  2. See quickstart
  3. Browse examples
  4. Join Discord