Eli Mernit

AWS and GCP are committing absolute robbery with their GPU prices

Introducing beam.cloud compute pools

> rent GPUs across 10+ regions
> H100s from 2.75/h
> 8xB300s available at this very moment
> Mount portable NVMe drives to any instance

You don't need to raise 100 million dollars to rent GPUs anymore - we're making compute available to everyone

most sandbox products are just interfaces over microVMs, almost like SSH or SMTP

everyone is treating the interface as the product, while the interface is more like an open protocol and the product is the runtime underneath

in 2016, the sandbox interface would have been a CNCF project - a standard interface for `vm new` and `vm stop` with swappable layers for Cloud Hypervisor, Firecracker, and QEMU

but because it's 2026 and everyone can ship code effortlessly, we get 100 companies trying to commercialize the same interface instead of having one open standard

PSA: beam.cloud is amazing for post-training

you can just train a bunch of models on cloud GPUs for pennies on the dollar

each model gets its own GPU sandbox that boots in half a second and shuts down afterwards

> bring your own Dockerfile
> run 100s of evals
> billed by the millisecond
> integrated into @harborframework

I trained 4 tiny models on 4 GPUs and it only cost $0.10

the biggest reason more people aren’t using cloud agents is because there’s not an obvious and simple way to migrate your local environment to the cloud

feels like there should be an open source tool from one of the coding harnesses that solves this

how fast is beam.cloud?

we're on a mission to be the most transparent cloud provider

not only are we open source, but we just released a comprehensive set of benchmarks on our system

some highlights:

> launch 100 sandboxes in 0.68 seconds
> launch 10,000 sandboxes in 12 seconds
> cold boot a sandbox with a custom 9.7Gi image in 3.7s

and you get all this starting at $0.00765/h

beam.cloud/benchmarks