Rent NVIDIA RTX 5090
Blackwell consumer GPU with 32 GB of GDDR7. Server sizes from 1 to 8 GPUs, billed by the hour. Pay in crypto, no KYC, root over SSH.
Live price by server size
Loading live data…| GPUs | Per GPU-hour | Server, per hour | Availability | Action |
|---|
Interruptible (spot)
| GPUs | Per GPU-hour | Server, per hour | Action |
|---|
Price history
Use the left and right arrow keys to read values at each point in time.
| Time (UTC) | Best 1-GPU hourly price | Free cards |
|---|
What you can run on the RTX 5090
The NVIDIA RTX 5090 is a Blackwell-generation consumer GPU with 32 GB of GDDR7 memory and hardware support for FP8 and FP4. It suits image and video generation, quantized language models and fine-tuning experiments. Rent it here by the hour, paid in crypto, with root access over SSH.
Typical tasks
- Image and video generation such as Stable Diffusion and Flux
- Inference of quantized 7-30B language models
- LoRA fine-tuning
- FP8 and FP4 inference, where your framework supports it
- Experiments on a budget
What fits
A rough inference guide, in billions of parameters: weights only, about 20% left for context; real limits depend on context length and the serving engine. It assumes the weights take up to 80% of the total memory, at 2 bytes per parameter for BF16, 1 byte for FP8 and half a byte for 4-bit. With several cards, the serving engine splits the model across them.
| GPUs | BF16 | FP8 | 4-bit |
|---|---|---|---|
| 1 GPU | 12B | 25B | 51B |
| 2 GPUs | 25B | 51B | 102B |
| 4 GPUs | 51B | 102B | 204B |
| 8 GPUs | 102B | 204B | 409B |
Based on 32 GB per card.
How to rent the RTX 5090
- Sign inSign in with your e-mail. No KYC: you pay from your own wallet.
- QuoteWe confirm the machine and the price within 1 hour, 03:00–23:00 UTC.
- PayYou approve a quote before anything is charged. Pay in USDC on Ethereum today; more tokens are coming.
- KeysAfter payment we test the server and send you SSH access.
Billed by the hour, no minimum term. Reserved terms for a month, a quarter or a year are quoted by hand and cost less per hour.
Open the order formBuilding for clients? See the partner program. AI agent? llms.txt describes the public price file and the order link.
RTX 5090 FAQ
What size of model fits on one RTX 5090?
As a rough guide for inference, with the weights taking up to 80% of the 32 GB: about 12B parameters at BF16, 25B at FP8 and 51B at 4-bit on a single card. Real limits depend on the context length and the serving engine. The What fits table on this page also covers 2, 4 and 8 cards.
How much memory does the RTX 5090 have?
32 GB of GDDR7 per card. The Specs section on this page shows the memory the cards on offer report right now.
Which low-precision formats does the RTX 5090 support?
FP8 and FP4 in hardware. Whether you can use them depends on your framework and on the model.
What can I run on the RTX 5090?
Typical uses are image and video generation, inference of quantized language models and LoRA fine-tuning. How large a model fits depends on its size and precision; the What fits table on this page gives a rough guide.
Do I need KYC?
No. You sign in with your e-mail and pay from your own wallet.