CoreWeave
Specialist CloudCoreWeave is a specialized cloud provider built from the ground up for GPU workloads. Offers some of the highest GPU density and fastest networking in the industry, with a focus on AI/ML training and inference at scale.
Cheapest On-Demand
$0.730/hr
Cheapest Spot
—
GPU Listings
20
Billing
On-demand, Reserved
Provider Info
Headquarters
Roseland, NJ
Founded
2017
Regions
US-East, US-West, EU-West
Min Commitment
None
Support
Standard → Enterprise
Strengths
- ▸Highest GPU density
- ▸InfiniBand networking
- ▸Kubernetes-native
- ▸Fast provisioning
Best For
Large-scale AI trainingLLM fine-tuningHigh-throughput inference
7-Day Price History · On-Demand $/hr
H100 80GBA100 80GBRTX 4090
Full GPU Catalog
| GPU Model | vRAM | On-Demand | Spot | Availability | Region |
|---|---|---|---|---|---|
| T4 | 16 GB | $0.730 | — | High | US-East |
| RTX 4090 | 24 GB | $0.890 | — | Med | US-West |
| RTX A5000 | 24 GB | $1.47 | — | Med | US-East |
| V100 32GB | 32 GB | $1.58 | — | Med | US-East |
| L4 | 24 GB | $1.88 | — | High | US-East |
| A30 | 24 GB | $2.12 | — | High | US-East |
| RTX A6000 | 48 GB | $2.83 | — | Med | US-East |
| A10G | 24 GB | $2.93 | — | High | US-East |
| A16 | 64 GB | $3.79 | — | Med | US-East |
| A40 | 48 GB | $4.24 | — | High | US-East |
| RTX 6000 Ada | 48 GB | $4.39 | — | Med | US-East |
| L40 | 48 GB | $4.42 | — | High | US-East |
| L40S | 48 GB | $5.88 | — | High | US-East |
| A100 80GB | 80 GB | $9.76 | — | High | US-East |
| H100 40GB | 40 GB | $14.74 | — | High | US-East |
| H100 80GB | 80 GB | $24.67 | — | High | US-East |
| H200 141GB | 141 GB | $42.48 | — | Low | US-East |
| B200 192GB | 192 GB | $56.93 | — | Low | US-East |
| GB200 NVL72 | 192 GB | $70.96 | — | Low | US-East |
| GB300 NVL72 | 288 GB | $96.40 | — | Low | US-East |