Skip to content
effecthub
Launched 2017 · Verified Oct 15, 2026

CoreWeave

Verified Oct 15, 2026 Active
Use tool
Share

GPU cloud provider renting Kubernetes-native compute by the hour for AI training and inference.

gpu cloudkubernetesai infrastructurePaid

Listed on EffectHub ·

Web · API
Screenshot of the CoreWeave website
#548
coreweave.com as we captured it — our own screenshot, not the vendor’s press image.

Pricing

Pricing model
Paid

On-demand instances are priced per hour, from $6.50/hour (GH200) up to $68.80/hour (8x HGX B200), with spot rates on some; GB300 NVL72, HGX B300 and one RTX PRO 6000 configuration show 'Contact sales'. Hourly billing has no matching unit here, so no headline price is set.

Use tool ↗
Overview

About CoreWeave

CoreWeave rents GPU compute, storage, and networking to companies training or running AI models, built on a Kubernetes-native platform rather than general-purpose cloud VMs. Pricing is published per GPU-hour and covers a range of NVIDIA hardware, from the GH200 at the low end to multi-GPU HGX B200 and GB200 NVL72 configurations at the high end; the newest chips show 'contact sales' instead of a rate. Founded in 2017 and headquartered in Livingston, New Jersey, CoreWeave has grown into a publicly traded infrastructure company whose customers include OpenAI, Mistral AI, and IBM. Access is through a web console and a Kubernetes-native API rather than a dashboard-only interface.

Platforms
Web · API
API
Yes
Last verified
Oct 15, 2026

Verified sources

  • Pricing page on-demand rates per hour: GH200 $6.50, L40 $10.00, L40S $18.00, A100 $21.60, GB200 NVL72 $42.00, HGX H100 $49.24, HGX H200 $50.44, HGX B200 $68.80; GB300 NVL72 and HGX B300 'Contact sales' (15 October 2026)coreweave.com ↗
  • Home page responds and presents CoreWeave as an AI cloud for training and inference (15 October 2026)coreweave.com ↗
Fact check

Is CoreWeave legit?

We don’t label tools safe or unsafe. Below are the records we checked ourselves, each with the date we read it.

Domain registered
Apr 1, 2019 (7 years ago)Registry record (RDAP), registrar GoDaddy.com, LLC, looked up Sep 6, 2026
TLS certificate
Issued by Let's Encrypt, valid until Jan 11, 2027Seen in our own connection to the site on Oct 15, 2026
Last checked
Oct 15, 2026The site answered
On the market since
2017The vendor's own material, read by us
Operated by
CoreWeave, Inc., United StatesOur research, card verified Oct 15, 2026

How we collect these records

Editorial verdict

Pros & cons of CoreWeave

high confidence

CoreWeave is a serious option for teams that need large GPU allocations and are comfortable with Kubernetes-native tooling rather than a simplified dashboard. Rates are published for on-demand use, but the largest and newest GPU configurations still require a sales conversation.

Pros

  • Publishes on-demand hourly rates instead of hiding all pricing behind sales
  • Offers a wide range of current NVIDIA GPU generations
  • Kubernetes-native platform integrates with standard orchestration tooling
  • Backed by large, named AI-lab customers

Cons

  • Newest and largest GPU configurations require contacting sales rather than self-serve pricing
  • Kubernetes-native operation assumes infrastructure expertise most small teams don't have

Best for

  • Teams training or serving large models who need Kubernetes-native GPU orchestration
  • Companies that have already standardized on Kubernetes for infrastructure

Not for

  • Individual developers wanting a simple pay-per-token API instead of infrastructure to manage
  • Small teams without Kubernetes experience who want a fully managed inference endpoint
#3inGPU Cloud & Compute· of 9 tools, ranked by rating and popularity

Be the first to review CoreWeave

Ask a question about CoreWeave