Cloudpods
cloudpods.org
- Free plan
- Open source
- Linux
- Web
- API
- Self-hosted
What Cloudpods is
Cloudpods is open-source software for managing enterprise IaaS and varied cloud infrastructure from one platform. It brings public and private cloud resources together through cloud accounts. Its KVM-based private cloud manages virtual machines, storage, networks, GPU cards, and bare metal. Documented connections include Alibaba Cloud, Azure, Tencent Cloud, AWS, Huawei Cloud, UCloud, Google Cloud, VMware, ZStack, DStack, and OpenStack. Cloudpods supports self-built Kubernetes clusters and management of existing clusters, including synced clusters from AWS EKS, Alibaba Cloud ACK, Azure AKS, and Tencent Cloud AKS. AI Cloud handles LLM inference and AI container applications with GPU scheduling, model libraries, templates, and images. Access includes multi-tenant IAM, multi-cloud SSO, the climc command-line tool, and REST APIs. Deployment options include ocboot for private cloud and Docker Compose or Kubernetes with Helm for multi-cloud. Open-source features are free; enterprise plans include annual options at 999 CNY a year and 1,999 CNY a year, plus a custom-priced edition.
Who it suits
Cloudpods suits teams managing a mix of public and private cloud resources, Kubernetes clusters, or AI workloads. It is also relevant to organizations that want a self-hosted option and can use its documented deployment approaches.
Good to know, and what to watch
- Manages public and private cloud resources together.
- KVM private cloud covers VMs, storage, and bare metal.
- Supports existing and self-built Kubernetes clusters.
- AI Cloud includes GPU scheduling and model libraries.
- Open-source features are free.
- Ollama inference and ComfyUI require a GPU.
- Advanced enterprise pricing depends on resources or customization.
- Paid enterprise editions include remote support.
Plans
Every plan and priceAdvanced Enterprise Edition
Open Source Community Edition
Basic Enterprise Edition — cloud management
Basic Enterprise Edition — private cloud
Free plan
The verdict
Cloudpods combines cloud management, Kubernetes support, and AI workload features, with open-source functionality available at no charge. Check GPU requirements for Ollama and ComfyUI, and compare the enterprise plan terms with your deployment needs.
What the maker publishes
- Access and APIs
- Cloudpods provides multi-tenant IAM, multi-cloud SSO, a command-line tool called climc, and REST APIs and models. source
- AI applications
- AI Cloud supports containerized applications including OpenClaw, Dify, ComfyUI and Hermes Agent. source
- AI features
- AI Cloud manages LLM inference and AI container applications, with GPU scheduling, model libraries, templates, and images. source
- AI inference
- AI Cloud can deploy Ollama, vLLM and SGLang inference services with GPU scheduling, model mounting and service-address allocation. source
- AI integrations
- The AI Cloud documentation lists Ollama, Dify, OpenClaw, HermesAgent, and ComfyUI as supported applications or inference services. source
- API surface
- Cloudpods exposes APIs through separate backend services, with internal Kubernetes endpoints and public NodePort endpoints. source
- Bare metal
- Cloudpods manages bare-metal resources and supports operating-system installation workflows. source
- Company headquarters
- The yunionio GitHub organization lists Beijing, China as its location. source
- Company identity
- Cloudpods is developed by Beijing YunionCloud Technology Co., Ltd., whose commercial YunionCloud product began development in 2017 and whose Cloudpods code was open-sourced on GitHub in 2019. source
- Deployment
- The documentation describes private-cloud deployment with ocboot and multi-cloud deployment through Docker Compose or Kubernetes with Helm. source
- Documentation depth
- The repository links separate getting-started, user, operations, development, API, Swagger, release-note and changelog documentation. source
- Identity and access
- The platform provides multi-cloud single sign-on, RBAC and IAM. source
- Install prerequisite
- The installer requires Ansible version 2.11.12 or later-compatible packages and Git. source
- Integrations
- Documented cloud platforms include Alibaba Cloud, Azure, Tencent Cloud, AWS, Huawei Cloud, UCloud, Google Cloud, VMware, ZStack, DStack, and OpenStack. source
- Kubernetes
- Cloudpods supports self-built Kubernetes clusters and management of existing clusters, including synced AWS EKS, Alibaba Cloud ACK, Azure AKS, and Tencent Cloud AKS clusters. source
- Kubernetes
- The feature set includes Kubernetes cluster management alongside database and infrastructure resources. source
- License
- The main Cloudpods repository is released under the Apache License 2.0. source
- Limit: deployment complexity
- The installer requires a clean operating system without preinstalled Kubernetes or Docker because it builds its own Kubernetes cluster and warns of conflicts otherwise. source
- Limit: open issues
- The public issue tracker shows 176 open issues, including reports about CPU display, ARM startup, container-port mapping, Ceph permissions and failed installations. source
- Limit: platform coverage
- The project’s own Ubuntu 22.04 support issue describes earlier deployment support as centered on CentOS 7, Debian 10 and Kylin, showing that OS compatibility has required separate adaptation work. source
- MCP integration
- An MCP server connects Cloudpods to clients such as Cursor and Cline through SSE or stdio. source
- MCP tools
- The MCP server documents 15 read and write tools for images, networks, regions, servers, storage, VPCs and monitoring. source
- Model library
- The model library manages model sources, versions and caches and supports reuse across instances and offline distribution. source
- Monitoring
- Monitoring and operations functions are included in the documented feature list. source
Checked against the maker's pages on 15 Oct 2026.