Aquanode

Aquanode gives you pods: a GPU box you build once and keep. An Environment (your image, everything you installed) and a Volume (your code, checkpoints, and data). Run it on any provider we support, or on hardware you already own, stop it, and start right where you left off.

The GPU is rented. The pod is yours.

Why that split matters

Renting a GPU is easy. Keeping what you built on it is not. Every provider hands you a raw machine, and the moment you give that machine back, the environment you spent an afternoon assembling goes with it. So people leave boxes running they are not using, or they rebuild from scratch every morning.

A pod separates the two. The machine is disposable and billed by the second, only while it runs. The pod is durable, costs $0.02/GiB per month to hold, and can come back on a different card, at a different provider, in a different country.

What you get

1. One account across every provider. No separate consoles, no separate billing, no separate access rules. See live rates and availability on the Marketplace.
2. Failover instead of errors. If the card you picked is taken while we are claiming it, we use another one that fits.
3. Environment versions. Share or keep one and it is numbered. Starting from it brings that exact install state back on a new machine.
4. Sharing. Send one Environment version as a link. Whoever opens it gets their own pod and their own empty Volume, your data stays untouched.
5. Auto-stop. Walk away and the pod saves itself, releases the machine, and stops the hourly charge.
6. Jobs. Run a saved pod version as GPU work that survives a provider taking the machine away, with a hard bound on what a run may cost.
7. A CLI that does all of it. aq covers the same surface.

Getting started