CoreWeave website screenshot

CoreWeave

CoreWeave is a specialized GPU cloud purpose-built for AI workloads, offering managed Kubernetes (CKS), Slurm-on-Kubernetes (SUNK), dedicated and serverless inference, AI Object Storage, distributed VAST file storage, HPC InfiniBand interconnect, and a Sandbox product. CoreWeave's control plane is Kubernetes-native and exposes APIs for CKS clusters, Inference deployments and gateways, VPCs, Object Storage, and Sandbox control.

5 APIs 7 Features
AICloudGPUHPCInferenceKubernetesMachine LearningStorage

APIs

CoreWeave Kubernetes Service API

The CKS API provisions and manages CoreWeave Kubernetes Service clusters and node pools on bare-metal GPU and CPU hardware. It exposes operations for cluster lifecycle, node poo...

CoreWeave Inference API

The CoreWeave Inference API manages Deployments, Gateways, and Capacity Claims for serverless and dedicated AI inference. It is used to create, update, and route to managed mode...

CoreWeave VPC API

The VPC API creates and manages Virtual Private Clouds on CoreWeave, including network configuration, routing, and isolation for CKS clusters and other compute resources.

CoreWeave AI Object Storage API

CoreWeave AI Object Storage (CAIOS) is an S3-compatible object storage service optimized for AI dataset and model storage. It supports standard S3 operations alongside CoreWeave...

CoreWeave Sandbox Control Plane API

The Sandbox Control Plane API provisions ephemeral compute sandboxes for short-lived, isolated workloads on CoreWeave infrastructure.

Collections

Pricing Plans

Coreweave Plans Pricing

1 plans

PLANS

Rate Limits

Coreweave Rate Limits

2 limits

RATE LIMITS

FinOps

Features

CoreWeave Kubernetes Service (CKS)

Managed Kubernetes on bare-metal GPU and CPU nodes for training, inference, and HPC.

SUNK

Slurm running on Kubernetes for batch and burst training workloads alongside K8s services.

CoreWeave Inference

Dedicated and serverless inference offerings with managed deployments, gateways, and capacity claims.

AI Object Storage

S3-compatible object storage purpose-built for AI dataset and model workloads.

Dedicated VAST Storage

High-performance VAST Data file storage for large-scale training pipelines.

HPC Interconnect

InfiniBand-based HPC fabric with GPUDirect RDMA for multi-node training.

CoreWeave Sandbox

Ephemeral compute sandboxes for short-lived, isolated workloads.

Integrations

Kubernetes

Native Kubernetes API surface across CKS clusters with standard kubectl and Helm workflows.

Slurm

Slurm workload manager integrated with Kubernetes through SUNK.

Terraform

Official CoreWeave Terraform provider for CKS clusters, VPCs, and object storage buckets.

NVIDIA GPU Operator

NVIDIA GPU Operator and InfiniBand fabric integration for accelerated workloads.

Resources

🔗
Website
Website
🔗
Developer
Developer
🔗
Documentation
Documentation
🌐
Portal
Portal
💰
Pricing
Pricing
📰
Blog
Blog
👥
GitHubOrganization
GitHubOrganization
🟢
StatusPage
StatusPage
📜
TermsOfService
TermsOfService
📜
PrivacyPolicy
PrivacyPolicy
💬
Support
Support
🔗
LinkedIn
LinkedIn
🔗
Terraform
Terraform
📦
SDKs
SDKs
🔗
Samples
Samples
🔗
LlmsText
LlmsText

Sources

apis.yml Raw ↑
opencollection: 1.0.0
info:
  name: CoreWeave Kubernetes Service (CKS) API
  version: v1beta1
request:
  auth:
    type: bearer
    token: '{{bearerToken}}'
items:
- info:
    name: Clusters
    type: folder
  items:
  - info:
      name: List all clusters
      type: http
    http:
      method: GET
      url: https://api.coreweave.com/v1beta1/cks/clusters
    docs: List all clusters
  - info:
      name: Create a cluster
      type: http
    http:
      method: POST
      url: https://api.coreweave.com/v1beta1/cks/clusters
      body:
        type: json
        data: '{}'
    docs: Create a cluster
  - info:
      name: Get a cluster's information by ID
      type: http
    http:
      method: GET
      url: https://api.coreweave.com/v1beta1/cks/clusters/:id
      params:
      - name: id
        value: ''
        type: path
        description: The cluster identifier.
    docs: Get a cluster's information by ID
  - info:
      name: Update a cluster's configuration by ID
      type: http
    http:
      method: PATCH
      url: https://api.coreweave.com/v1beta1/cks/clusters/:id
      params:
      - name: id
        value: ''
        type: path
        description: The cluster identifier.
      body:
        type: json
        data: '{}'
    docs: Update a cluster's configuration by ID
  - info:
      name: Delete a cluster by ID
      type: http
    http:
      method: DELETE
      url: https://api.coreweave.com/v1beta1/cks/clusters/:id
      params:
      - name: id
        value: ''
        type: path
        description: The cluster identifier.
    docs: Delete a cluster by ID
bundled: true