Anyscale Pricing
Maximize GPU utilization with Ray on Anyscale.
Anyscale offers you a pay-as-you-go approach. Only pay for the compute you use on demand.
Unlock volume discounts and other benefits that scales with your usage. Have GPU reservations? Use them. The more you run, the more you save.
| Deploy how you want | Hosted Fastest way to get started with your first project. Fully managed infrastructure with no setup required. | Bring Your Own Cloud (BYOC) Ready for Ray production workloads. Deploy inside your own cloud, or on-prem. Expert support available 24x7. |
|---|---|---|
| Deployment |
|
|
| Data Residency |
|
|
| Compute |
|
|
| Billing |
|
|
| Support |
|
|
Only pay for what you use, no monthly fixed fees. Unlocks discounts as usage grows.
Compute Costs
Instance Containing:
CPU Only
from $0.0112 /hr
NVIDIA T4
from $0.1264 /hr
NVIDIA L4
from $0.1966 /hr
NVIDIA A10G
from $0.3545 /hr
NVIDIA A100
from $0.6388 /hr
NVIDIA H100
from $1.8603 /hr
NVIDIA H200
from $2.1411 /hr
An end-to-end workload with data processing, batch inference, distributed training and online serving.
Fine-tune an LLM to perform batch inference and online serving for entity recognition.
Build and deploy MCP servers using Ray Serve in HTTP streaming and STDIO modes.