What's good about...
CoreWeave
- Access to latest NVIDIA GPU generations
- Kubernetes-native cloud platform for easy orchestration
- Transparent pricing with up to 60% savings on reserved capacity
- Purpose-built for AI and ML workloads
Vast.ai
- Large pool of over 20,000 GPUs
- Over 68 GPU types across 40+ data centers
- Enterprise-grade security with SOC2 certification
- Pre-built templates including PyTorch, CUDA, and TensorFlow
Price comparison
Vast.ai Pricing
Vast.ai operates as a global marketplace where pricing is billed per second and broken into three components:
-
Active rental cost for the GPU, charged only when an instance is running.
-
Storage cost for allocated disk space, which applies as long as the instance exists (even when stopped).
-
Bandwidth costs for data transfer.
GPU Fleet Overview
Starting Price
CoreWeave
Vast.ai
Billing Options
CoreWeave
Vast.ai
Available Models
CoreWeave offers 10 GPU models, while Vast.ai offers 55.
Heads up: We do our best to keep these specs & prices accurate. However, cloud costs may fluctuate based on region, usage, and other factors not listed here. These are estimates based on common setups and are for informational purposes only. Always verify current rates & exact specs with the provider before provisioning.
Which services do they offer
Here are some managed services that CoreWeave and Vast.ai offer:
Company details
CoreWeave
|
Vast.ai
|
|
|---|---|---|
| Website | coreweave.com | vast.ai |
| Headquarters | United States of America | United States of America |
| Founded | 2017 | 2018 |
| Data Center Locations | 13 | -- |
| Example Customers | Jane Street, OpenAI, MistralAI, NovelAI | IBM, Brave, Speechify, Cognition, Bosch |
CoreWeave
Vast.ai
CoreWeave
Vast.ai
CoreWeave
Vast.ai
CoreWeave
Vast.ai
CoreWeave
Vast.ai
Alternatives to consider
Compare CoreWeave and Vast.ai against other cloud providers.
Our data for CoreWeave was last updated on June 11, 2026, and for Vast.ai on June 11, 2026.
H100