Keep Anyscale.
Change where it runs.
Anyscale is the managed platform for Ray, built by the people who created it. heata can be added as a compute provider inside your Anyscale account, so you keep the whole Anyscale experience — workspaces, job queues, services, observability — while the work itself runs on compute that heats homes.
Pricing
Base compute plus the data science platform tier, which covers distributed computing APIs such as Ray and Spark. Your Anyscale licence is separate and billed by Anyscale.
Component | Unit | Price |
|---|---|---|
vCPU (managed) | core-hour | £0.017 |
RAM (managed) | GB-hour | £0.0025 |
Data science platform tier | CPU-hour | 0.71p |
Data science platform tier | GB-RAM-hour | 0.09p |
GPU (RTX 4090) | GPU-hour | +£0.89 |
Egress | TB | £20.00 |
Prices exclude VAT and are billed monthly in arrears. As a worked example, a 4 CPU / 8 GB instance costs about 4p per hour in platform fees on top of base compute. Reserved commitments take 30% off base compute for one year, or 55% for three.
Talk to us
Point Anyscale at compute that heats homes
A short setup on our side and your Ray workloads are providing hot water. Tell us about your account.