dstack

Visit site Alternatives

Score7.4

Rank#1 of 21

PriceFree

Free planYes

Runs onAPI, Linux, macOS, Self-hosted, Web, Windows

Summary

dstack is an open-source orchestration layer for running AI workloads across GPU clouds, Kubernetes, virtual machines, and bare-metal clusters. It provisions infrastructure and schedules jobs, with support for auto-scaling, port forwarding, and ingress. YAML configurations define fleets, development environments, tasks, services, presets, and volumes. Supported accelerators include NVIDIA, AMD, TPU, and Tenstorrent; documented backends include AWS, Azure, GCP, Kubernetes, GPU cloud providers, remote SSH hosts, and experimental Slurm. Tasks can use frameworks such as accelerate, torchrun, Ray, and Spark. The CLI and HTTP API provide ways to manage resources and connect integrations, and the server can run on a laptop or another environment that can access the clusters. Services can expose inference endpoints through gateways with HTTPS, custom domains, auto-scaling, and rate limits. The self-hosted open-source plan is free. Server data and project secrets are plaintext by default, unless administrators configure AES-256-GCM encryption. TPU support is limited to single-host instances of up to eight cores.

Who it is for

dstack suits teams orchestrating AI workloads across different accelerator and infrastructure environments. It is also suited to users who want to configure workloads with YAML and manage resources through a CLI or HTTP API.

What is good

Free, self-hosted open-source plan.
Supports NVIDIA, AMD, TPU, and Tenstorrent accelerators.
Works with Kubernetes and cloud backends.
Inference gateways offer HTTPS and auto-scaling.
CLI and HTTP API are available.

What to know first

Server data is plaintext by default.
Secrets are plaintext unless server encryption is configured.
TPUs are limited to single-host instances of up to eight cores.
GPU Marketplace usage requires prepaid credits.

Verdict

dstack brings workload configuration, infrastructure provisioning, and job scheduling together across varied compute environments. Consider its default plaintext storage and single-host TPU limit when evaluating it.

dstack plans and pricing

All plans

dstack OSS Free Open-source orchestration stack; self-hosted dstack.ai · 30 Sept 2026

dstack Sky GPU Marketplace Not published GPU usage is pay-as-you-go; prices are per GPU-hour and vary by GPU and provider. On-demand and spot GPU compute; listed GPU-hour price ranges; prepaid credits dstack.ai · 30 Sept 2026

Compared on GPU cluster management software

Free plan: Yesdstack.ai
Deployment model: hybriddstack.ai
Workload scheduling: bothdstack.ai
Kubernetes support: Yesdstack.ai
Quota controls: Nodstack.ai
GPU utilization metrics: Yesdstack.ai
Cloud GPU support: Yesdstack.ai

Facts

What it does: dstack is an open-source orchestration layer for AI workloads across GPU clouds, Kubernetes, VMs, and bare-metal clusters.dstack.ai · 30 Sept 2026
Workload types: It supports fleets, development environments, tasks, services, presets, and volumes configured with YAML files.dstack.ai · 30 Sept 2026
Accelerators: dstack supports NVIDIA, AMD, TPU, and Tenstorrent accelerators out of the box.dstack.ai · 30 Sept 2026
Inference: Services can deploy model inference as endpoints, and gateways support HTTPS, auto-scaling, custom domains, and rate limits.dstack.ai · 30 Sept 2026
Integrations: Documented backends include AWS, Azure, GCP, Kubernetes, multiple GPU cloud providers, remote SSH hosts, and an experimental Slurm backend.dstack.ai · 30 Sept 2026
Framework compatibility: The maker describes dstack as compatible with any hardware, open-source tools, and frameworks.dstack.ai · 30 Sept 2026
Interfaces: Users can manage resources with the dstack CLI or call its HTTP API.dstack.ai · 30 Sept 2026
Platforms: The CLI runs on Linux, macOS, and Windows; the server can be installed on those systems, with Windows using WSL 2.dstack.ai · 30 Sept 2026
Deployment: The server can run on a laptop or another environment with access to the cloud and on-prem clusters being used.dstack.ai · 30 Sept 2026
Security: Server data is stored in plaintext by default; administrators can configure AES-256-GCM encryption for stored data.dstack.ai · 30 Sept 2026
Secrets: Secrets are project-scoped, managed by project admins, and stored in plaintext by default unless server encryption is configured.dstack.ai · 30 Sept 2026
Support: The documentation directs users to report issues on GitHub and ask questions in the dstack Discord server.dstack.ai · 30 Sept 2026
Cost model: dstack Sky does not currently charge for BYOC mode; GPU Marketplace usage is prepaid and resource prices are shown in the console before provisioning.dstack.ai · 30 Sept 2026
Commercial offering: dstack Factory extends the open-source product with advanced multi-tenancy, usage metering, billing automation, and optimized inference presets for frontier open models.dstack.ai · 30 Sept 2026
Purpose: dstack is an open-source orchestration layer for AI workloads on heterogeneous accelerators, including GPU clouds, Kubernetes, VMs, and bare-metal clusters.dstack.ai · 30 Sept 2026
Workloads: It supports fleets, dev environments, tasks, services, experimental presets, and volumes through YAML configurations.dstack.ai · 30 Sept 2026
Provisioning: dstack manages infrastructure provisioning and job scheduling, including auto-scaling, port forwarding, and ingress.dstack.ai · 30 Sept 2026
Integrations: Listed backend types include AWS, Azure, GCP, Kubernetes, Slurm, Runpod, Nebius, Lambda, and other cloud providers.dstack.ai · 30 Sept 2026
Frameworks: The tasks guide names accelerate, torchrun, Ray, and Spark as distributed frameworks that work with dstack.dstack.ai · 30 Sept 2026
API: dstack offers an HTTP API for functionality not available in the CLI and for integrations that need to call the server directly.dstack.ai · 30 Sept 2026
Service endpoints: Services can be published with HTTPS, custom domains, auto-scaling, and rate limits through gateways.dstack.ai · 30 Sept 2026
Security: The service configuration supports authorization, which is enabled by default for services.dstack.ai · 30 Sept 2026
Deployment limit: The TPU guide says dstack currently supports single-host TPUs only, with a maximum of eight cores per TPU instance.dstack.ai · 30 Sept 2026
Support: The dstack Sky page directs users with questions or needing help to Discord or to contact the company directly.dstack.ai · 30 Sept 2026
Hosted pricing: dstack Sky Marketplace pricing is dynamic by provider, shown in the console before provisioning, and billed against prepaid credits.dstack.ai · 30 Sept 2026
Company: The terms identify dstack Inc. as a Delaware corporation with offices in Dover, Delaware, United States.dstack.ai · 30 Sept 2026