CoreWeave Release Notes

Default storage quotas for CoreWeave AI Object Storage and Distributed File Storage have been reduced.

The default AI Object Storage quota is now 20 TiB per availability zone, and the default DFS quota is now 10 TiB per CKS cluster.

See AI Object Storage quotas and DFS quotas for current values.

You can now configure CoreWeave Telemetry Relay yourself, without contacting Support. Use the Cloud Console or the Telemetry Relay API to forward CoreWeave audit logs to an HTTPS endpoint. See the Telemetry Relay self-service release notes for more information.

A new NodePools and Cluster Autoscaler dashboard is now available in CoreWeave Grafana.

The dashboard gives you a single place to inspect Node Pool capacity and conditions, read Node Pool events, and track Cluster Autoscaler scaling activity, errors, and logs without querying metrics or logs directly.

You can now browse, upload, download, rename, and delete objects in CoreWeave AI Object Storage buckets directly from the Cloud Console, with no CLI tools or third-party software required. Navigate folder hierarchies, search for buckets and objects by name, and drag files from your desktop to upload them. See the release notes for the full list of supported operations and required permissions.

CKS Node Pools now support targetRacks for NVL72-powered instances, letting you request GB200 and GB300 capacity at the rack level instead of counting individual Nodes. Each rack provisions a full set of 18 Nodes, so you no longer need to set targetNodes to a multiple of 18 to deploy complete racks.

CKS Node Pools now support prefill, which provisions a replacement Node before a triaged Node is drained and removed.

With prefill enabled, the pool stays at or above spec.targetNodes during maintenance, so workloads do not see a capacity gap while the replacement is delivered.

See Enable Node Pool Prefill to configure the timeout and maxNodes fields.

New AI Object Storage documentation covers bucket lifecycle policies, aborting incomplete multipart uploads, and emptying and deleting buckets. The S3 reference now consolidates API compatibility notes, and the Terraform guide documents a known issue with NoncurrentVersionTransitions in CoreWeave provider v0.7.0. See the release notes for more information.

The supported GPU drivers have changed. See Supported driver versions for the full compatibility table.

GB200 NVL72-powered instances

Drivers removed: 570

A100

Drivers removed: 570

H100 (InfiniBand)

Drivers removed: 570

H200 (InfiniBand)

Drivers removed: 570

L40

Drivers removed: 570

L40S

Drivers removed: 570

RTX Pro 6000 Blackwell Server Edition

Drivers removed: 570

The supported GPU drivers have changed. See Supported driver versions for the full compatibility table.

B200 (InfiniBand)

New default: 580 → 595

B300 (InfiniBand)

Drivers added: 595

GB200 NVL72-powered instances

New default: 580 → 595

GH200

New default: 580 → 595

H100 (InfiniBand)

New default: 580 → 595

H200 (InfiniBand)

New default: 580 → 595

L40

New default: 580 → 595

L40S

New default: 580 → 595

Drivers added: 535, 570, 595

The Cabinet Wrangler and Cabinet Visualizer dashboards now display rack name as the primary label for filtering and identification, replacing NVLink domain. Both metrics remain available in the dashboards. See the Cabinet Wrangler release note for more information.

Version 1.18.0 of the CoreWeave cert-manager Helm chart switches the bundled Let's Encrypt ClusterIssuers from HTTP01 to DNS01 challenges, resolved through a CoreWeave webhook at acme.coreweave.com. An ingress controller is no longer required for certificate issuance, and wildcard certificates are now supported. See the cert-manager DNS01 release notes for more information.

CoreWeave AI Object Storage now supports conditional writes. Attach HTTP precondition headers (If-None-Match or If-Match) to PutObject, CompleteMultipartUpload, and CopyObject requests to make writes atomic and prevent accidental overwrites without client-side locking. See the conditional writes release notes for more information.

CoreWeave Omni is now available. CoreWeave Omni is a cloud-as-a-service model in which CoreWeave deploys and operates the full CoreWeave cloud stack inside your data center. You retain ownership of the facility and hardware while CoreWeave delivers a managed region, the CoreWeave Cloud Platform, and day-to-day operational management. For availability, sizing, and pricing, contact your CoreWeave account team. See the CoreWeave Omni release notes for more information.