Features
What Aquanode actually does, capability by capability. Every one of these is shipped and in the product today.
On-Demand GPUs
Deploy H100, H200, A100, and B200 GPUs instantly, billed per second, with no minimum commitment.
Multi-Provider Marketplace
One live marketplace over every provider we support, so you can compare real prices for the same GPU before you deploy.
Workspace Backups & Migration
Your Environment and Volume are captured every time you Stop a pod, and again on a background tick while it runs. Restore either onto a different GPU, or a different provider entirely.
Browser Terminal
A real root shell on the box from a console tab, with no SSH client, on providers that give the box a dialable address.
VM Templates
JupyterLab or ComfyUI running and reachable on boot with credentials generated for you, plus plain OS images.
Startup Scripts
Save an init script once and attach it to any deployment. It runs as root on every fresh boot.
Volumes
Your /workspace: code, checkpoints, and datasets, with automatic history and no manual save button. Attaches to any pod, on any provider.
SSH Keys
Register a public key once; it is injected into every VM you deploy, and the console shows which instances each key unlocks.
API Keys
Scoped, expiring team tokens for CI/CD and your own tooling, against the same API the console uses.
Teams & Workspaces
Shared workspaces where deployments, volumes, API keys and the credit balance belong to the team, not to one login.
Live GPU Metrics
Watch GPU utilization, VRAM, temperature and power on a running deployment, refreshed every few seconds.
Portable Workspaces
A named directory that follows a VM deployment across providers. Attach it, sync it, detach it, attach it to the next box.
Stop & Start
Stop a GPU pod to stop the bill, then start it on the same provider or a different one, with your Environment and Volume back the way you left them.
Auto-stop
A GPU you walked away from gets a warning, then its environment and volume are captured and the machine released. Opt-in per pod.
aq CLI
Rent a GPU, provision it, and get a shell in one command from your laptop.
Guides by workload
Running a specific workload? See the guides for ComfyUI, fine-tuning, Jupyter notebooks, multi-GPU nodes, ready-made pods and teams. Keeping your work between sessions is covered by Volumes, Stop & Start, Auto-stop and workspace backups.
