RunKeyRunKey

Buy credits once.
Spend them by the second.

No spot market, no tiers, no subscription. Every GPU has a fixed credit rate per second, and you spend credits only while an instance is actually running.

600 credits

Basic pack

$9.99one-time

$0.0167 per credit

  • ≈13 hours on an RTX A4000, ≈5.7 on a 4090
  • Any GPU model, at its own credit rate
  • Root on the real GPU over Direct SSH
  • CUDA and cuDNN, pre-installed
  • 80 GB disk and unmetered traffic
  • Metered by the second, no minimum
  • Credits never expire
Buy 600 credits
Most Popular50% OFF

3,600 credits

Pro pack

$29.99one-time

$0.0083 per credit

  • ≈78 hours on an RTX A4000, ≈34 on a 4090
  • Any GPU model, at its own credit rate
  • Root on the real GPU over Direct SSH
  • CUDA and cuDNN, pre-installed
  • 80 GB disk and unmetered traffic
  • Metered by the second, no minimum
  • Credits never expire
Buy 3,600 credits
60% OFF

9,000 credits

Max pack

$59.99one-time

$0.0067 per credit

  • ≈196 hours on an RTX A4000, ≈85 on a 4090
  • Any GPU model, at its own credit rate
  • Root on the real GPU over Direct SSH
  • CUDA and cuDNN, pre-installed
  • 80 GB disk and unmetered traffic
  • Metered by the second, no minimum
  • Credits never expire
Buy 9,000 credits

Credits never expire. Purchases are non-refundable.

RTX A4000

16 GB · 448 GB/s · 19.2 TFLOPS FP16

0.0127 credits/sec
45.71 credits per hour

The efficient workhorse. Ideal for SD 1.5 and SDXL at moderate batch sizes, LoRA training and always-on inference where cost per hour matters more than raw speed.

  • ›SD 1.5 / SDXL
  • ›LoRA training
  • ›Long-running inference
Deploy RTX A4000

RTX 3090

24 GB · 936 GB/s · 35.6 TFLOPS FP16

0.0198 credits/sec
71.43 credits per hour

24 GB of VRAM at the lowest price on the platform. A dependable choice for SDXL, Flux at reduced precision, and 13B-class model inference.

  • ›SDXL at scale
  • ›Flux (quantised)
  • ›13B inference
Deploy RTX 3090

RTX 4090

24 GB · 1.01 TB/s · 82.6 TFLOPS FP16

0.0294 credits/sec
105.71 credits per hour

The default pick. Ada-generation throughput makes Flux, SDXL and video pipelines roughly twice as fast as a 3090 for well under twice the price.

  • ›Flux at full precision
  • ›Video pipelines
  • ›Fast iteration
Deploy RTX 4090

RTX 5090

32 GB · 1.79 TB/s · 104.8 TFLOPS FP16

0.0393 credits/sec
141.43 credits per hour

Blackwell, 32 GB, and the fastest memory bandwidth we rent. For long video generation, high-resolution batches and workflows that will not fit in 24 GB.

  • ›Long video generation
  • ›High-res batches
  • ›30B+ inference
Deploy RTX 5090

Multi-GPU instances spend linearly — a 4× RTX 4090 instance is four times the single-card rate.

Included in every instance

The credit rate is the whole bill. There is no second line.

  • 80 GB of disk on every instance
  • Root access over Direct SSH
  • CUDA and cuDNN pre-installed, matched to the node's driver
  • Unmetered inbound and outbound traffic
  • Per-second metering with no minimum
  • Automatic termination at zero balance
How you pay

Buy a credit pack, and instances draw it down as they run. Credits never expire, and there is no monthly fee whether you run one instance or twenty.

Pricing questions

What is a credit?
The unit everything on the platform is billed in. You buy credits in a pack, and a running instance spends them by the second at a rate fixed per GPU model — an RTX A4000 costs 0.0127 credits a second, an RTX 4090 costs 0.0294. Nothing else consumes credits.
Is the rate really fixed?
Yes. An RTX 4090 costs the same credits per second whichever node it lands on and whenever you deploy it. We hold that rate and absorb the variation upstream, so a number you worked out once stays right.
How is a multi-GPU instance billed?
Linearly. An instance with four RTX 4090s spends four times the single-card rate. There is no volume discount and no multi-GPU premium — the arithmetic is the arithmetic.
Do credits expire?
No. Purchased credits stay on the account until you spend them, however long that takes. There is no subscription and no monthly minimum.
Do I pay for storage or bandwidth?
No. Each instance includes 80 GB of disk and unmetered traffic in its credit rate. When the instance is terminated its disk goes with it, so there is nothing left behind to bill you for.
Am I charged while an instance is starting?
No. The meter starts when the instance reports running and its connection details appear. Time spent scheduling and pulling the image is on us, and an instance that never comes up is never billed.
What if I forget to terminate something?
The console shows the burn rate and how many hours of runway are left, and warns you as it gets short. If the balance reaches zero the instance is terminated automatically — the account cannot quietly run up a bill.
What is the refund policy?
Purchases are non-refundable, and credits do not expire — they are there whenever you come back to them.

Start with the Basic pack

600 credits is about thirteen hours on an A4000. Spend them whenever you like.

Create an account