# Hanzo Cloud — the AI-native cloud > The AI-native cloud: models and agents, functions, containers, sandboxes, machines, GPUs and Kubernetes, every data primitive an AI application needs, networking, identity and one trace end to end. One API, one organization, one usage ledger — from code to production without assembling the cloud first. Run it managed on Hanzo Cloud, attach your own account or cluster, or self-host the same open-source image. Hanzo Cloud # The AI‑native cloud. Run models, agents, applications, data and infrastructure through one operating plane. Kubernetes-native. Fully observable. Open source. Run it on Hanzo infrastructure or infrastructure you control. [Start building](https://console.hanzo.ai)[Explore the cloud](https://hanzo.ai/products) GPUs · Functions · Containers · Databases · Observability · One API $ docker run -p 8080:8080 ghcr.io/hanzoai/cloud ## One cloud for the whole AI application. Not a model API beside a container host beside a database beside a tracing vendor. One system, with the parts already introduced to each other. - ModelInference, embeddings, routing, fallbacks. - AgentTools, memory, sandboxes, orchestration. - ApplicationFunctions, containers, machines, endpoints. - DataRelational, vector, objects, cache. - InfrastructureClusters, networking, regions, capacity. - OperationsIdentity, secrets, traces, billing. Everything shares the same organization, project, identity, policy, usage and operational context. ## One runtime for every kind of work. The same plane, held for a millisecond or held for a month. What changes is how much of a machine the work needs, not what you have to learn. - FunctionsScale to zero between calls. - ContainersYour image, deployed and addressable. - SandboxesIsolated shells for agents and untrusted code. - MachinesA whole machine, by the second. - GPUsAccelerators attached to any of them. - KubernetesThe substrate all of it runs on. ContainersDeploy an image and invoke it. Deployments, triggers, secrets, metrics and logs are all part of the same surface, so a function you ship is a function you can watch.api.hanzo.ai/v1/functions Machines Sandboxes ## The cloud is bigger than one provider. One control plane, different infrastructure underneath it. The API, the console, the identities and the bill do not change when the machines do. ### Hanzo Cloud Capacity we operate and meter, provisioned by the call. ### Your own cloud account Your credentials, your machines, the same control plane over them. ### Your own cluster Attach Kubernetes you already run and address it like any other. Machines, GPUs, clusters and Kubernetes are one surface whichever of those it lands on. Visor is the runtime underneath them, and it meters every provider against the org that asked. [Compute across your own accounts](https://hanzo.ai/visor) ## Every model. One cloud interface. Let Enso choose when you want the best answer for the job. Name a model when you want exactly that one, every time. Run open weights when you want to hold them yourself. Whichever you pick, auth, budgets, routing, fallbacks, usage and observability stay in one system — so changing your mind about a model is a string, not a migration. OpenAI-compatible·Anthropic-compatible [Every model we serve](https://hanzo.ai/models)[Build and run agents](https://hanzo.ai/agents) ## The data primitives AI applications actually use. Not one database with an extension bolted on. Each primitive is the right shape for the job, and they share a project, a key and a bill. ### Vector Embeddings stored and searched where they are used. ### SQL Relational, per project and per tenant. ### KV Cache and queues at request latency. ### Object Storage Objects, signed URLs, lifecycle rules. ### Datastore Columnar analytics over everything you collected. ### Base An application backend — records, rules, realtime. ### DocDB Documents on the MongoDB wire protocol. ### Memory What an agent is allowed to remember between runs. ## Everything connects without becoming your problem. Services find each other. Traffic reaches them over TLS on a name you chose. The edges are closed until you open one. The gateway fronts every API, DNS answers for your names, and routing between services is something the cloud does rather than something you write. [The networking plane](https://hanzo.ai/network) ## See the entire system, not separate products. One request, from the edge through an agent, a model, a tool, a database and a function — and what every hop of it cost. One trace. End to end. Logs, metrics, traces, evaluations and spend are read from the same request, so “why was that answer slow” and “why was that answer expensive” are the same question with the same answer. [Tracing, metrics and evaluation](https://hanzo.ai/o11y)[Live status, in public](https://status.hanzo.ai) ## Security is part of the cloud. Every person, agent and workload has an identity. Permissions follow the resource rather than the caller. Secrets stay scoped to the environment that needs them. Actions stay attributable to whoever took them. None of that is a product you add afterwards — an agent gets an identity because it runs here, and the trail is written because the call went through the same door as everything else. [How identity, keys and audit work](https://hanzo.ai/security) ## One API. Every capability. - api.hanzo.ai/v1/models - api.hanzo.ai/v1/agents - api.hanzo.ai/v1/functions - api.hanzo.ai/v1/sql - api.hanzo.ai/v1/vector - api.hanzo.ai/v1/s3 - api.hanzo.ai/v1/… Same key. Same organization. Same project. Same policy model. Same usage ledger. The interface changes. The system doesn’t. Twelve operations, one key, no glue. Each is checked against the published OpenAPI document on every build, and a beat the document does not carry is dropped rather than drawn. One org, one afternoon: a spaceship shop that did not exist this morning. `GET /v1/models`Returns the list of available models from the routing table. Models Pause ## Push code. Get a production system. No separate build service to wire up. No container registry to stand up. No secret product to buy. No logging to set up afterwards. What comes back is a running system with an address, a certificate, its secrets, its metrics, its logs and its share of the bill — because those were never separate purchases. [Builds, releases and rollback](https://hanzo.ai/registry)[The CLI](https://hanzo.ai/cli) ## Your cloud shouldn’t be a one-way door. The platform is open source, and the image we operate is the image you can pull. Same API, same console, same data — on your cluster, in your network, or on a laptop on a plane. $ curl -fsSL https://hanzo.sh | sh $ hanzo serve cloud [What is open, and under which license](https://hanzo.ai/open-source)[View source](https://github.com/hanzoai) ## Pay for what runs. Nothing is billed for existing. A function that was not called and a machine that was not running cost what they did — nothing. Models by usage Compute by execution Storage by consumption Managed services by usage One organization balance and one usage ledger behind all of it, so every line traces back to a request somebody made. [Every plan and every rate](https://hanzo.ai/pricing) ## Explore the cloud. Every plane, and what is in it. ### AI [Models](https://hanzo.ai/models) · [Providers](https://hanzo.ai/cloud/providers) · [Agents](https://hanzo.ai/agents) · [Inference](https://hanzo.ai/cloud/inference) · [Embeddings](https://hanzo.ai/cloud/embeddings) ### Compute [GPUs](https://hanzo.ai/cloud/gpus) · [Machines](https://hanzo.ai/machines) · [Functions](https://hanzo.ai/functions) · [Edge](https://hanzo.ai/edge) · [Applications](https://hanzo.ai/cloud/applications) ### Data [Vector](https://hanzo.ai/vector) · [SQL](https://hanzo.ai/sql) · [KV](https://hanzo.ai/kv) · [Object Storage](https://hanzo.ai/s3) · [Datastore](https://hanzo.ai/datastore) ### Dev [CLI](https://hanzo.ai/cli) · [SDKs](https://hanzo.ai/sdks) · [API](https://hanzo.ai/api) · [Integrations](https://hanzo.ai/integrations) · [IDE](https://hanzo.ai/cloud/ide) ### Observe [Overview](https://hanzo.ai/overview) · [Evals](https://hanzo.ai/cloud/evals) · [Logs](https://hanzo.ai/cloud/logs) · [Metrics](https://hanzo.ai/metrics) · [Traces](https://hanzo.ai/o11y) ### Security [IAM](https://hanzo.ai/iam) · [Authz](https://hanzo.ai/authz) · [KMS](https://hanzo.ai/kms) · [Secrets](https://hanzo.ai/cloud/secrets) · [Audit](https://hanzo.ai/cloud/audit) ### Network [Gateway](https://hanzo.ai/gateway) · [Nodes](https://hanzo.ai/cloud/nodes) · [VPC](https://hanzo.ai/cloud/vpc) · [DNS](https://hanzo.ai/dns) · [Load Balancer](https://hanzo.ai/cloud/load-balancer) [View the complete catalog](https://hanzo.ai/products) ## Build on a cloud designed for AI. Models. Agents. Compute. Data. Security. Observability. One API from prototype to production. [Start building](https://console.hanzo.ai)[Read the docs](https://docs.hanzo.ai/docs/services/cloud) Or run Hanzo Cloud on infrastructure you control.