SkyForge · pricing
Pay by the second. Stop, and you stop paying.
One per-second meter across every GPU we rent — no minimums, no tier gates, no reserved commitment to get started. One SkyForge API key across every instance, live spend in the console, terminate anytime.
per-second billing · no minimums · one key, every instance · terminate anytime
01On-demand
The rate card.
Every GPU type in the catalog, priced per GPU-hour with our platform fee already included — the same rates the console launches from, billed per second from ready to terminate.
| GPU | VRAM / GPU | GPUs per node | From |
|---|---|---|---|
| A4000 | 16 GB | 1–4 | $0.28 / GPU-hr |
| RTX 4090 | 24 GB | 8 | $0.52 / GPU-hr |
| A16 | 16 GB | 2–4 | $0.67 / GPU-hr |
| RTX A6000 | 48 GB | 1–8 | $0.74 / GPU-hr |
| RTX 5090 | 32 GB | 8 | $0.91 / GPU-hr |
| L40S | 48 GB | 1, 2, 4, 8, 10 | $1.15 / GPU-hr |
| L4 | 24 GB | 1–4 | $1.24 / GPU-hr |
| RTX 6000 Ada | 48 GB | 1–4 | $1.26 / GPU-hr |
| L40 | 48 GB | 1–8 | $1.29 / GPU-hr |
| GAUDI2 | 96 GB | 8 | $1.43 / GPU-hr |
| A100 80 GB | 80 GB | 1–8 | $1.76 / GPU-hr |
| A10 | 24 GB | 1 | $1.79 / GPU-hr |
| A100 40 GB | 40 GB | 1, 8 | $1.82 / GPU-hr |
| RTX A5000 | 24 GB | 1–4 | $1.85 / GPU-hr |
| A40 | 48 GB | 1 | $2.42 / GPU-hr |
| RTX PRO 6000 | 96 GB | 1, 8 | $2.85 / GPU-hr |
| V100_32G | 32 GB | 1–4 | $3.04 / GPU-hr |
| V100 | 16 GB | 1 | $3.04 / GPU-hr |
| H100 | 80 GB | 1–2 | $3.25 / GPU-hr |
| H200 | 141 GB | 8 | $5.20 / GPU-hr |
“From” is the lowest available per-GPU-hour rate for each type; multi-GPU nodes scale linearly. Rates include our platform fee and are billed per second while the instance is running. Live availability is in the console.
02How the meter works
Simple, honest, and yours to control.
No hourly rounding, no plans to graduate into, and nothing running that you didn't ask for. The meter tracks real usage — and the dashboard shows it live.
Pay by the second
The meter runs per second while an instance is up — not rounded to the hour. You pay for exactly the time you use.
Stop and you stop paying
Terminate an instance and the meter stops. There's nothing left running and nothing left to pay for.
No minimums, no tier gates
No monthly commitment and no plan to upgrade into. Every capability is available from your first instance.
One API key, every instance
A single SkyForge API key works across all of your instances and OpenAI-compatible endpoints — no per-instance keys to juggle.
Live spend in the dashboard
See accrued, burn rate, and projected spend update live in the console, so the bill is never a surprise.
No bandwidth or egress fees
Network transfer is unmetered. Serve as much traffic as your instance can handle — the only meter is time.
Terminate anytime
Spin up and tear down whenever you like. No lock-in, no cancellation flow — the instance is yours to end on your terms.
03Reserved & committed
Steady workload? Reserved is a conversation.
Committed-use pricing and dedicated capacity for sustained production, quoted to your usage and allocated through our team — including reserved Blackwell B200 / B300. There's no public rate card; tell us the workload and we'll size it.
04FAQ
Questions about billing.
05Get started
Launch now, or talk to us about reserved.
On-demand is self-serve from the console — no sales call. Running steady production? Tell us what you're building and we'll size committed capacity.