dstackai/dstack
dstack is an open-source control plane for running development, training, and inference jobs on GPUs—across hyperscalers, neoclouds, or on-prem.
This tool helps machine learning engineers and researchers manage and run their GPU-intensive development, training, and inference workloads across various cloud providers or on-premises hardware. It takes your code and configurations (defining resources, tasks, and services) and deploys them, handling the complexities of provisioning and orchestrating GPU compute resources. The ideal user is someone who needs to efficiently utilize GPUs for AI/ML projects.
2,062 stars. Actively maintained with 72 commits in the last 30 days. Available on PyPI.
Use this if you are a machine learning engineer or researcher juggling GPU resources across multiple cloud platforms or on-prem systems and want a unified way to manage your model development, training runs, and deployment.
Not ideal if you primarily work with CPU-only workloads or have minimal needs for distributed GPU computing.
Stars
2,062
Forks
217
Language
Python
License
MPL-2.0
Category
Last pushed
Mar 13, 2026
Commits (30d)
72
Dependencies
26
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/mlops/dstackai/dstack"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related tools
skypilot-org/skypilot
Run, manage, and scale AI workloads on any AI infrastructure. Use one system to access & manage...
ray-project/kuberay
A toolkit to run Ray applications on Kubernetes
kubeflow/kale
Kubeflow’s superfood for Data Scientists
volcano-sh/volcano
A Cloud Native Batch System (Project under CNCF)
m3dev/gokart
Gokart solves reproducibility, task dependencies, constraints of good code, and ease of use for...