An open-source system to run, manage, and scale AI workloads across Kubernetes, Slurm, and 20+ clouds or GPU providers with one interface.
yunionio/cloudpods
Open-source cloud-native platform that unifies on-premise KVM, bare metal, vSphere and many public clouds behind one set of APIs.
Open-source alternative to:VMware vCenter
AWS Management Console
- Stars
- 2,962
- Forks
- 637
- Open issues
- 203
- Last commit
- 3 hours ago
What is it?
Cloudpods is a multi- and hybrid-cloud management platform written in Go. It lets an organization manage on-premise KVM hosts, bare-metal servers, and VMware vSphere clusters alongside resources from many cloud accounts, hiding provider differences behind a single set of APIs.
The supported providers include AWS, Azure, Google Cloud, Alibaba Cloud, Huawei Cloud, Tencent Cloud and several other Chinese public clouds, as well as OpenStack, Proxmox VE and Nutanix. Managed resources cover compute instances, disks, networks, VPCs, load balancers, object storage, RDS, Elastic Cache, DNS and NAS.
The project also includes an AI Cloud module for deploying and scheduling LLM inference services such as Ollama, vLLM and SGLang, plus AI container applications, with GPU detection and model library management.
Key features
- 01Manages on-premise KVM, bare-metal and vSphere resources alongside public cloud accounts.
- 02Exposes one unified API across AWS, Azure, Alibaba Cloud and many other providers.
- 03Provides VPC, load balancer, object storage, RDS, DNS and NAS resource management.
- 04Deploys and schedules LLM inference services with GPU support.
Use cases
- Turning a few physical servers into a private cloud with automated bare-metal lifecycle management.
- Adding a self-service cloud layer on top of an existing VMware vSphere cluster.
- Giving one console access to several public cloud accounts in a multi-cloud setup.
- Running LLM inference and AI container workloads on a shared GPU pool.
Quick start
- 1
Read the Getting Started guide and pick the installation scenario that fits you
- 2
Follow the official documentation to install and run Cloudpods
Things to know
- Full deployment targets private or hybrid cloud infrastructure, so it is heavy for simply trying the project out.
- The README links to external documentation for installation, and no single package-manager install command is given.
Topics
FAQ
- What is cloudpods?
- Open-source cloud-native platform that unifies on-premise KVM, bare metal, vSphere and many public clouds behind one set of APIs. Cloudpods is a multi- and hybrid-cloud management platform written in Go. It lets an organization manage on-premise KVM hosts, bare-metal servers, and VMware vSphere clusters alongside resources from many cloud accounts, hiding provider differences behind a single set of APIs.
- Is cloudpods free to use? What license does it have?
- cloudpods is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
- Is cloudpods still maintained?
- Actively maintained. The last commit to cloudpods was on Oct 21, 2026.
- What can I use cloudpods for?
- Turning a few physical servers into a private cloud with automated bare-metal lifecycle management. Adding a self-service cloud layer on top of an existing VMware vSphere cluster. Giving one console access to several public cloud accounts in a multi-cloud setup. Running LLM inference and AI container workloads on a shared GPU pool.
- How do I install cloudpods?
- Read the Getting Started guide and pick the installation scenario that fits you Follow the official documentation to install and run Cloudpods
- Where can I get cloudpods?
- The source code and installation instructions are on GitHub: https://github.com/yunionio/cloudpods
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A Python framework for building, scaling and deploying AI and ML workflows, from notebook prototypes to production orchestrators.
Open-source real-time infrastructure monitoring agent with per-second metrics, auto-discovery, built-in ML anomaly detection and alerting.
A Go terminal UI for real-time log analysis, with live charts, severity breakdowns, regex filtering, and optional AI-powered insights.
Kong is a cloud-native, platform-agnostic API, LLM and MCP gateway, extensible through plugins, for routing and securing service traffic.
A PyTorch-based distributed training and inference system for large AI models, offering parallelism, memory optimization and heterogeneous training.