Vibeleaderboard
Index / tool
Visit www.coreweave.com
Category
Developer Tools
Rank
No. 1368Tools index

Previous survey · No. 1327 ·

Pricing
Paid
Type
TOOL
Builder
@CoreWeave
Date

About

Large-scale GPU infrastructure. Kubernetes-native GPU cloud built for large training and inference fleets. On-demand and reserved capacity across current-generation accelerators.

What it does

The site's pricing page lists hourly rental rates for a wide range of NVIDIA GPU servers, alongside CPU instances, object and file storage tiers, and networking add-ons like dedicated connections, all sold as pay-per-hour cloud compute.

Stated on the product site

Pricing model
Compute is billed hourly with separate on-demand and spot rates, plus volume discounts for committed usage.
Platform
A managed Kubernetes control plane is included as part of the offering, listed as free.
Storage tiers
Object storage is sold in multiple tiers priced per gigabyte per month.
Networking
Basic networking features such as NAT gateway and VPC are provided free, while public IPs and dedicated direct connections carry separate listed fees.
Inference pricing
A separate per-GPU hourly inference rate is listed alongside standard on-demand and spot prices for several GPU models.

Not stated on the site

  • The page does not describe the interface used to provision or manage instances, such as a web console, CLI, or API.
  • It's not stated on this page what industries or company sizes the service primarily targets.
  • Regions available beyond North America and Europe are not listed on the page.
  • Contract terms or minimum commitment lengths for reserved capacity discounts are not shown.

Written from the product site at coreweave.com.

Tags

CoreWeaveGPU Cloud

Comments (0)

No comments yet

Editorially curated, with community endorsements as a secondary signal. Corrections welcome.