Create clusters — POST /v1/compute/clusters

Attaches a BYO cluster to the caller's org — the kubeconfig is validated, KMS-sealed and added to the fleet — and answers 201 with the cluster as it now appears on GET /v1/compute/clusters.

POST /v1/compute/clusters

Compute → · All API references → · Get started →

Address	`https://api.hanzo.ai/v1/compute/clusters`
Method	`POST`
Operation	`attachCluster`
Auth	`Authorization: Bearer $HANZO_API_KEY`

Attaches a BYO cluster to the caller's org — the kubeconfig is validated, KMS-sealed and added to the fleet — and answers 201 with the cluster as it now appears on GET /v1/compute/clusters. Billed the nominal management fee: the customer brings the compute, Hanzo meters the management plane.

Request

4 fields, body application/json (required).

Field	In	Type	Required	Description
`default`	body	boolean	—	Default marks this the org's default cluster for scheduling.
`kubeconfig`	body	string	—	Kubeconfig is the cluster's kubeconfig, verbatim.
`name`	body	string	—	Name is the fleet-local name for the cluster; lower-cased, and the key the detach route addresses it by.
`provider`	body	string	—	Provider is a free-form label for where the cluster runs ("gke", "on-prem"); it is display only, not a routing key.

Response

Status	Body	Meaning
`200`	compute.clusterView	ok
`default`	problem-details	refused

200 body — 19 fields.

Field	In	Type	Always	Description
`amdGpu`	body	integer (int64)	—	AmdGPU is the same count for `amd.com/gpu`: AMD accelerators across the BYO cluster's nodes, as of the attach.
`createdAt`	body	string	—	CreatedAt is when the cluster started existing: the earliest creation time among its pools for a managed cluster, and for a BYO one the RFC 3339 moment it was attached.
`doClusterId`	body	string	—	DoClusterID carries the SAME id as DoksClusterID.
`doksClusterId`	body	string	—	DoksClusterID is the provider's own id for the cluster, and the value the /v1/compute/k8s/clusters/:id routes take.
`kind`	body	string	—	Kind says which of the two kinds of cluster this row is, and there are only two: "managed" — Visor provisioned it and Hanzo's account pays the provider — or "byo", an existing cluster the org attached by kubeconfig.
`name`	body	string	—	Name is the cluster's name: the provider's for a managed cluster, and for a BYO one the lower-cased fleet name it was attached under — which is also how the detach route addresses it.
`nodeCount`	body	integer (int64)	—	NodeCount is how many worker nodes the cluster has — the sum over its pools for a managed cluster, and for a BYO one the node count read off the cluster when it was attached.
`nodePools`	body	compute.nodePoolView[]	—	NodePools is the authoritative node inventory — every pool, each with its own size and count.
`nodePools[].autoScale`	body	boolean	—	AutoScale reports whether the provider's cluster autoscaler owns this pool's size, moving Count between MinNodes and MaxNodes as workloads demand.
`nodePools[].count`	body	integer (int64)	—	Count is how many nodes the pool has right now.
`nodePools[].maxNodes`	body	integer (int64)	—	MaxNodes is the ceiling the autoscaler will not grow the pool past, and so the bound on what this pool can cost.
`nodePools[].minNodes`	body	integer (int64)	—	MinNodes is the floor the autoscaler will not shrink the pool below.
`nodePools[].name`	body	string	—	Name is the pool's name as the provider knows it.
`nodePools[].poolId`	body	string	—	PoolID is the provider's id for the pool — the value the scale and delete routes address it by.
`nodePools[].size`	body	string	—	Size is the provider size slug every node in the pool runs at ("s-4vcpu-8gb", "gpu-h100x8-640gb").
`nodeSize`	body	string	—	NodeSize is a display convenience: the size slug of the FIRST pool.
`nvidiaGpu`	body	integer (int64)	—	NvidiaGPU is how many NVIDIA accelerators the cluster's nodes advertise, the sum of `nvidia.com/gpu` allocatable across them.
`region`	body	string	—	Region is the provider region slug for a managed cluster.
`status`	body	string	—	Status is the cluster's state: the provider's own word for a managed cluster ("running", "provisioning"), "unknown" when the provider stated none, and always "attached" for a BYO cluster — that one says the kubeconfig is on file, not that the cluster is reachable this second.

Failure carries the platform error shape — see Errors.

Examples

CLI

SDK

HTTP

MCP

hanzo compute clusters create

.ts

.py

.go

.rs

.java

import { Configuration, ComputeApi } from 'hanzoai';

const api = new ComputeApi(new Configuration({ accessToken: process.env.HANZO_API_KEY }));
const { data } = await api.attachCluster({ default: false, kubeconfig: "<kubeconfig>" });

from hanzoai.cloud import ApiClient, Configuration
from hanzoai.cloud.api import ComputeApi

client = ApiClient(Configuration(access_token=os.environ["HANZO_API_KEY"]))
result = ComputeApi(client).attach_cluster(default=False, kubeconfig="<kubeconfig>")

cfg := hanzoai.NewConfiguration()
cfg.AddDefaultHeader("Authorization", "Bearer "+os.Getenv("HANZO_API_KEY"))
client := hanzoai.NewAPIClient(cfg)

resp, _, err := client.ComputeAPI.AttachCluster(context.Background()).Execute()
if err != nil {
	return err
}

use hanzo_client::apis::{configuration::Configuration, compute_api};

let mut cfg = Configuration::new();
cfg.bearer_access_token = std::env::var("HANZO_API_KEY").ok();

let result = compute_api::attach_cluster(&cfg, Default::default()).await?;

import ai.hanzo.cloud.ApiClient;
import ai.hanzo.cloud.api.ComputeApi;

ApiClient client = new ApiClient();
client.setBearerToken(System.getenv("HANZO_API_KEY"));

var result = new ComputeApi(client).attachCluster();

curl -X POST https://api.hanzo.ai/v1/compute/clusters \
  -H "Authorization: Bearer $HANZO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
       "default": false,
       "kubeconfig": "<kubeconfig>"
     }'

Tool compute, op attachCluster — POST the JSON-RPC envelope to https://api.hanzo.ai/v1/mcp.

curl -X POST https://api.hanzo.ai/v1/mcp \
  -H "Authorization: Bearer $HANZO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
       "jsonrpc": "2.0",
       "id": 1,
       "method": "tools/call",
       "params": {
         "name": "compute",
         "arguments": {
           "op": "attachCluster",
           "input": {
             "default": false,
             "kubeconfig": "<kubeconfig>"
           }
         }
       }
     }'

Compute API · All Hanzo APIs · Interactive reference

Was this page useful?

PreviousCreate verifications NextReplace agent