What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Cloud computing is on-demand access over a network to a shared pool of configurable computing resources—such as servers, storage, networks, applications, and services—that can be provisioned and released with little management effort. Today, cloud platforms also offer managed components for building and operating AI agents. Those agent platforms extend the cloud model; they do not replace its underlying infrastructure or definition.
What does cloud computing mean?
The National Institute of Standards and Technology (NIST) published its formal definition in 2011. It describes cloud computing as a way to access shared, configurable computing resources on demand, rather than as a synonym for anything hosted remotely or reachable through the internet. NIST’s definition is intended as a framework for describing cloud offerings, not as a ranking of providers. Read NIST SP 800-145.
For a service to fit the framework, it is considered against five essential characteristics. They describe how resources are made available and managed:
- On-demand self-service: A customer can provision capabilities such as server time or storage as needed, without a provider employee handling each request.
- Broad network access: Services are available over a network through standard mechanisms used by different kinds of clients.
- Resource pooling: Provider resources serve multiple customers, with physical and virtual resources dynamically assigned and reassigned.
- Rapid elasticity: Resources can expand or contract with demand, often automatically.
- Measured service: Use is metered at an appropriate level so it can be monitored, controlled, and reported.
These are the five characteristics in NIST’s 2011 framework, not a measure of cloud adoption or market size. NIST SP 800-145 full text.
#1 Best Overall
How does cloud computing work?
A cloud service combines physical computing resources with software that abstracts and manages them. NIST describes a physical layer of hardware—typically servers, storage, and network components—and an abstraction layer of software deployed over that hardware. The abstraction lets customers request configurable services without having to manage each underlying physical component themselves.
- A client requests a service over a network. This might be an application, storage, or computing capacity.
- Provider software allocates resources. It presents capacity as configurable services and manages how physical or virtual resources are assigned.
- Hardware performs the work. Servers, storage, and networking support the requested service.
- Usage is measured. Metering supports monitoring, control, and reporting.
The details differ by product: cloud customers do not necessarily see where physical resources are located or how a service is implemented. NIST notes that resource pooling commonly gives customers location independence, while permitting location to be specified at a higher level, such as a country, state, or data center. NIST SP 800-145.
Remote hosting or internet access alone does not establish that a service meets the cloud definition. NIST’s service-evaluation guidance can help assess whether a capability aligns with the framework and which service model best describes it. NIST’s 2018 evaluation guidance.
What are IaaS, PaaS, and SaaS?
Infrastructure as a Service, Platform as a Service, and Software as a Service are service models. They distinguish what the provider supplies and operates from what the customer deploys or configures.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errors| Model | What the provider supplies | What the customer does |
|---|---|---|
| IaaS Infrastructure as a Service |
Fundamental computing, storage, and networking resources. | Runs software on those resources. |
| PaaS Platform as a Service |
Provider-supported tools and runtime environments for deploying applications. | Deploys applications using the platform. |
| SaaS Software as a Service |
A provider-run application accessed through a client, such as a browser. | Uses the application. |
This is a responsibility distinction, not a scale from “less cloud” to “more cloud.” A service’s precise boundaries depend on the offering; NIST’s framework supplies the broad categories. NIST SP 800-145.
What do public, private, community, and hybrid cloud mean?
Public, private, community, and hybrid cloud are deployment models. They describe how cloud infrastructure is provisioned for and shared among an organization or group, and whether distinct cloud infrastructures are connected. They answer a different question from IaaS, PaaS, and SaaS, which describe the service being supplied.
Rank #3
| Deployment model | What it describes in NIST’s framework |
|---|---|
| Private cloud | Cloud infrastructure provisioned for a single organization. |
| Community cloud | Cloud infrastructure provisioned for a community of organizations with shared concerns. |
| Public cloud | Cloud infrastructure provisioned for open use by the general public. |
| Hybrid cloud | Distinct cloud infrastructures connected so they can support a combined arrangement. |
These are NIST’s deployment categories; they do not, by themselves, specify a particular provider, technology, or level of security. NIST SP 800-145.
How are cloud platforms expanding to support AI agents?
Cloud platforms are adding managed services for AI agents: software systems that can use models and tools to carry out multi-step tasks. These offerings build on cloud infrastructure and add components for development, runtime, connections to other systems, identity and permissions, state, governance, and observability. “Agentic cloud” describes this direction in provider architectures; it is not a new formal category in NIST’s cloud definition.
Development and operations
Google Cloud documents a managed agent lifecycle covering development, runtime, security, governance, and observability. Its documented development paths include a visual low-code environment, a managed Agents API, and a code-first Agent Development Kit. These are Google Cloud’s available approaches as described in its documentation, not universal requirements for building agents. Google Cloud agents overview.
Connecting agents to enterprise systems
A Google Cloud reference architecture shows one way to assemble an agent system: an orchestrator agent runs on Cloud Run and coordinates work across enterprise systems; Model Context Protocol (MCP) servers expose backend systems through standardized tools; and state can be stored in agent sessions or Cloud Storage. The design recommends least-privilege IAM service accounts, authentication controls, structured logs and traces, and infrastructure as code for repeatable deployments. It is an example architecture, not a prescription for every workload. Google Cloud’s agentic AI reference architecture.
Managed agent runtimes
AWS announced the general availability of Amazon Bedrock AgentCore on October 13, 2025, describing it as a managed platform for building, deploying, and operating agents with connectivity, runtime, security, and monitoring capabilities. In a September 18, 2026 article, AWS describes AgentCore Runtime as a managed compute layer and discusses its support for longer-running autonomous workloads. These are AWS’s descriptions of its services, not independent performance tests or evidence of industry-wide adoption. AWS announcement of AgentCore general availability; AWS article on AgentCore Runtime.
The underlying needs have not disappeared: agents still depend on compute, networking, storage, identity, and operational controls. What is changing is the layer above those foundations, with managed components that can host models and agents, connect tools and data, preserve state, control permissions, and help operators observe behavior. The implementations cited here are provider examples, not a neutral provider comparison. Google Cloud reference architecture; Google Cloud agents overview; AWS AgentCore announcement.
Recommended Free Tools
How should you evaluate a cloud approach?
Start with the workload and its constraints rather than asking which cloud is “best” in the abstract. NIST’s service and deployment models help describe the arrangement; practical requirements determine whether it fits. For agent workloads, the reference architectures also make identity, integration, state, governance, and observability important design considerations.
- Service responsibility: Decide whether the workload needs raw infrastructure, a managed application platform, or a provider-run application, and identify what the customer still operates.
- Deployment arrangement: Determine whether public, private, community, or connected hybrid infrastructure fits the organization’s arrangement.
- Workload and reliability needs: Define what the workload must do and what level of operational control it requires; the cited sources do not provide a neutral provider-by-provider scorecard.
- Data location: Establish any country, state, or data-center location requirements, then check whether the specific offering supports them.
- Identity and permissions: For agents that can invoke tools or access enterprise systems, determine how authentication and least-privilege access will be applied.
- Interoperability: Check how the service connects to the required tools and systems, including whether its interfaces and protocols fit the existing environment.
- Governance and observability: Establish how activity, state, and behavior will be monitored and governed.
- Operational effort and cost: Compare the work and usage model required by the specific services under consideration; the framework itself does not establish which option will cost less.
For a candidate service, evaluate the actual offering against NIST’s characteristics and service models rather than relying on its “cloud” label alone. NIST’s evaluation guidance.
Quick Recap
What to remember
- Cloud computing means on-demand network access to a shared pool of configurable resources, not simply remote hosting.
- NIST’s 2011 framework sets out five essential characteristics, three service models, and four deployment models; the service and deployment models classify different things.
- Agent platforms add managed tools for running and operating agents on top of cloud foundations. Provider architectures illustrate possible designs, not universal standards or independent proof that one platform is best.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
