---
title: "Create a cluster"
method: POST
path: "/v1/orgs/{orgId}/clusters/create"
tags: ["Clusters"]
---

# Create a cluster

`POST /v1/orgs/{orgId}/clusters/create`

## Path parameters

- `orgId` string, required

## Request body

- CreateClusterRequest
  - `allowFallback` boolean
  - `containerCommand` string[]
  - `containerPort` integer
  - `diskSizeGb` integer
  - `envVars` object
  - `gpuModelId` string
  - `gpusPerReplica` integer
  - `healthCheckPath` string
  - `healthCheckPort` integer
  - `healthChecks` string[]
  - `imageUrl` string
  - `kernelModules` string[]
  - `name` string, required
  - `preferredRegions` string[]
  - `privileged` boolean
  - `registryCredentialId` string
  - `replicaCount` integer
  - `resourceSize` string
  - `templateId` string
  - `tier` string

## Response `200`

Created

- ClusterSummary
  - `activeReplicas` integer
  - `createdAt` string
  - `endpointUrl` string
  - `gpuModelId` string
  - `id` string, required
  - `name` string, required
  - `organizationId` string, required
  - `replicaCount` integer
  - `status` string, required

## Other responses

- `400` — The request is invalid
- `401` — Missing or invalid API key
- `403` — API key lacks the required scope
- `409` — The request conflicts with existing state
- `503` — The request is valid but the capacity it needs is momentarily unavailable: a transient saturation that clears on node or VM turnover, not a durable conflict. Codes: NODE_AT_CAPACITY, INSUFFICIENT_GPU_CAPACITY, INSUFFICIENT_CPU_CAPACITY, INSUFFICIENT_IP_CAPACITY. Retry the same request with exponential backoff; the Retry-After header suggests how long to wait.

---

[API](https://skmtc.net/openrelay/apis/openrelay-api.md) · [All operations](https://skmtc.net/openrelay/apis/openrelay-api/llms.txt) · [OpenAPI document](https://skmtc-service-staging.skmtc.workers.dev/v1/apis/openrelay/openrelay-api/revisions/ed76b41e8006/schema)