Renters
Rent GPUs on lium.io — browse the marketplace, deploy a containerized pod in minutes, pay only for what you use.
This section is UI-first: every page walks through the dashboard at lium.io and shows where to click. The matching API and CLI live in the Developers section and are referenced from each page in a collapsed "For agents and automation" block.
Get going
| Step | What you do | Page |
|---|---|---|
| 1 | Sign up, top up, paste an SSH key | Quickstart |
| 1b | The whole first hour, timed: top-up, CLI, node choice, GPU check, result file, stop billing | Your first hour |
| 2 | Pick a machine, configure a pod, click Deploy | Create a pod |
| 3 | Check the GPU interconnect and the download link before a multi-GPU job | Before you rent 8 GPUs |
| 4 | Pick the right Docker image (PyTorch, custom, …) | Templates |
| 5 | Run one job across several machines over InfiniBand or RoCE | Multi-node clusters |
| 6 | Keep your data across pod sessions | Volumes |
| 7 | Schedule automatic backups of your work | Backups |
| 8 | Restore a backup into a new or existing pod | Restores |
| 9 | Auto-terminate pods so they stop charging | Scheduled termination |
| 10 | Know what you pay: per-second charges, the 5-minute cycle, running out of balance | Billing |
| 10b | What one pod cost, day by day, kept after the pod is removed | Pod statements |
| 11 | Open a private Discord channel with the provider for a rented pod | Discord support |
| 11 | Put your company name, address and VAT/tax id on the invoice for card payments | Invoices |
Reference (don't read top-to-bottom)
- Broken pods — what a BROKEN status means (provider force-closed the pod) and how to clear it.
- If your pod disappears — why a pod can leave your list on its own, where the record is, and why every result is hashed and pushed off the pod as it is produced.
- Pod overview & security model — what a pod actually is and what the GPU provider can/can't see.
- API Keys — generate keys for the CLI, SDK, MCP, or your own AI agent.
- Pod security — what to never put on a non-CVM pod.
- GPU profiling with ncu — rent a whole-host machine where Nsight Compute works.