CoreWeave
coreweave.com
Specialized GPU cloud provider for AI workloads.
Quick Information
- Category:
- MLOps & Infrastructure
- Pricing:
- Contact Sales
- Platforms:
- Cloud, Kubernetes
- Free Plan:
- No
- API Available:
- Yes
- Last Verified:
- October 2026
What Is CoreWeave?
CoreWeave is a specialised GPU cloud provider for AI workloads. Its list price for an H100 is about $6.16 per GPU-hour, but the reserved rate is poorly documented and the sources conflict. Most say the published rate is per 8-GPU node, at about $49.24 an hour, or $6.16 per GPU-hour: an NVIDIA HGX H100 8-GPU node lists at $49.24 per node-hour on demand as of August 2026, though one guide says the same H100 is still listed on a legacy Classic pricing page at $4.76 an hour. On spot, the 8x H100 node drops from $49.24 to $19.71 an hour, about $2.46 per GPU. CoreWeave says reserved capacity can be up to 60% cheaper but does not publish the terms, with third-party estimates varying from roughly $2.23 an hour for a two-year contract to about $1.49 to $2.10 an hour for three-year contracts on an older page. The 60% discount maths gives about $2.46 per GPU-hour, but that is an estimate.
For comparison, the same sources list Crusoe H100 at $3.90, Lambda at $3.99 and RunPod Community at $2.69 per GPU-hour, so CoreWeave's list price is higher, reflecting its networking and enterprise focus. I did not reach CoreWeave's own pricing page, so these figures come from third-party aggregators dated August to October 2026, and you should confirm current rates and reserved terms with CoreWeave directly, along with storage and networking charges that are not covered here.
CoreWeave suits large AI labs and enterprises that need big, high-performance clusters with reserved capacity. Compare Lambda, Crusoe and Nebius. Browse the MLOps & Infrastructure tools directory for more.
What Can You Use CoreWeave For?
Large-scale GPU training
HGX clusters.
Enterprise AI infrastructure
Reserved capacity.
CoreWeave Features
- Node-based pricing — 8 GPUs per node.
- Spot discounts — About 60% off.
- Unpublished reserved terms — Negotiated.
Who Is CoreWeave For?
- AI labs — Scale.
- Enterprises — Reserved capacity.
CoreWeave Pros & Limitations
What We Like
- Large high-performance clusters
- Spot discounts
- Reserved options
Limitations
- Higher list price per GPU
- Reserved terms unpublished
- Pricing from third parties
CoreWeave Pricing
| Plan | Price | Main Features |
|---|---|---|
| H100 on demand | $49.24 per 8-GPU node-hour (about $6.16 per GPU-hour) | As of August 2026; legacy Classic page $4.76 |
| H100 spot | $19.71 per 8-GPU node-hour (about $2.46 per GPU) | Reported |
| Reserved capacity | Up to 60% off (terms unpublished); estimates $1.49-$2.46 per GPU-hour | By contract |
Pricing was last verified October 2026. Check the official website for the latest pricing.
AI Listing Jungle Ranking
- Global Rank:
- #577
- Overall Score:
- 62.5/100
- Monthly Users:
- –
- Rank Change:
- ▲ 983
This ranking reflects AI Listing Jungle's own independent scoring methodology and is not an endorsement by CoreWeave or its company. How Our Rankings Work
Frequently Asked Questions
How much is an H100 on CoreWeave?
About $6.16 per GPU-hour on demand.
Is spot cheaper?
About $2.46 per GPU-hour.
Are reserved prices public?
No; up to 60% off is claimed.
How does it compare?
Higher than Lambda, Crusoe and RunPod list prices.
Our Take
CoreWeave is built for big AI workloads.
Negotiate reserved terms.
CoreWeave Alternatives
CoreWeave Comparisons
Ready to try CoreWeave?
Visit the official website to get started.
Visit Official WebsiteAI Listing Jungle is an independent AI tools directory. Product names, logos, and trademarks belong to their respective owners.