Questions, answered.

At CryptGPU. Dedicated NVIDIA H100 SXM servers (80 GB HBM3, NVLink 4) with 1, 2, 4 or 8 GPUs cost $1,779 per GPU per month, paid in Bitcoin, Ethereum, USDT, Monero or Litecoin. There is no KYC, and the server is online in under 10 minutes once the payment is confirmed on-chain. See the H100 server prices.

CryptGPU accepts Monero for every server it rents, from one NVIDIA RTX 5080 at $199 a month to eight NVIDIA B300, with no KYC. A Monero payment keeps the sender, the receiver and the amount private. See how to pay with Monero.

Yes. CryptGPU runs no KYC: no ID documents, no selfie and no identity verification, whatever the GPU or the amount. You pay in crypto, and the server details are sent to your email address. See what we ask and never ask.

CryptGPU prices every GPU at least 30% below the median public on-demand price of 40 GPU clouds: $4,019 for the NVIDIA B300, $3,439 for the B200 and $2,279 for the H200, per GPU per month, from 1 to 8 GPUs. Each of these prices is also below the lowest on-demand list price we found on 23 September 2026. See every price.

In FP8, a 70B model needs about 84 GB of GPU memory: the cheapest CryptGPU server that holds it is one NVIDIA RTX PRO 6000 (96 GB) at $959 a month. In 4-bit it needs about 46 GB: a server with two NVIDIA RTX 5090 (64 GB in total) at $658 a month. Long contexts and many parallel users need more. See the GPU memory of each open model.

When it runs most of the day. A CryptGPU H100 at $1,779 a month costs less than an H100 at the median on-demand price once it runs more than about 17 hours a day. For occasional use, an hourly cloud costs less. See the break-even for every GPU.

Yes. When you order at CryptGPU, the Software step can install vLLM, SGLang or Ollama with Open WebUI and the model you pick (Qwen, DeepSeek, gpt-oss, Gemma, GLM, Kimi, Mistral and more), ComfyUI or JupyterLab, at no extra cost. See the open models and the servers they need.

CryptGPU accepts USDT on Tron (TRC-20) and on Ethereum (ERC-20), as well as Bitcoin, Ethereum, Monero and Litecoin. Prices are in US dollars and USDT follows the dollar, so the amount barely moves between the quote and the payment. See how to pay with USDT.

Rent a GPU server with enough memory for an open model, then run Ollama with Open WebUI for a chat interface, or vLLM for an OpenAI-compatible API, bound to localhost and reached over SSH. At CryptGPU, the Software step can preinstall either with the model you pick, at no extra cost. See how to self-host an LLM and the Ollama guide.

CryptGPU runs no KYC and asks only for an email address and a crypto payment; paying with Monero keeps the payment private on the blockchain. Anonymous does not mean invisible: we keep order records and see connection metadata, and the law and our Acceptable Use Policy apply. See how to rent a GPU server anonymously.

Every CryptGPU configuration has one fixed price per month. You pay the month in advance, your server runs for that month, and you renew for the next one if you want to keep it. There is no hourly metering.

Hourly clouds must price in idle hours, short rentals and constant churn. Renting by the month removes most of that cost, and we pass it on: each GPU is priced 30% below the market median, rounded down.

For each GPU we collected the public on-demand list prices of 40 GPU clouds on 23 September 2026, took the median hourly price per GPU and multiplied it by 730 hours. Spot and marketplace offers, and AWS, Google Cloud and Azure, are excluded. See every number.

Yes. A server with n GPUs costs n times the per-GPU price, from 1 to 8 GPUs. vCPU, RAM and NVMe scale in the same proportion.

One NVIDIA RTX 5080 at $199 per month, with 8 vCPU, 32 GB of DDR5 RAM and 0.5 TB of NVMe.

No. The price is fixed for the paid month. If our list prices change, the new price applies only to months paid after the change.

No. Monthly is cheaper when your GPUs run most of the day: above about 17 hours a day against the median hourly price. For occasional use, an hourly cloud costs less. Compare for your GPU.

You can pay with BTC, ETH, USDT, XMR and LTC. One payment covers one month of your server.

The invoice shows the amount in the cryptocurrency you choose, for a limited quote window. Send the exact amount, on the network shown on the invoice, within that window.

You do. The full invoice amount must arrive, so add your wallet’s network fee on top.

Once it is confirmed on-chain. Your server starts at that point.

No. Payments are in cryptocurrency only: BTC, ETH, USDT, XMR or LTC.

Contact us with the invoice and the transaction ID before sending anything else.

No. The GPUs in your server are dedicated to you for the whole month: no time-slicing and no other tenant on your cards.

Data-center GPUs (B300, B200, MI355X, H200, H100) and the RTX PRO 6000 and L40S come in servers of 1, 2, 4 or 8. GeForce cards (RTX 5090, 4090, 5080) come in servers of 1, 2 or 4.

Each GPU comes with a fixed amount of vCPU, DDR5 RAM and local NVMe, which scales with the number of GPUs. The per-GPU amounts are listed on the pricing page and on each GPU page.

No. Each server has one GPU model. Configure two servers to run two models.

Start from memory: your model and its context must fit in GPU memory. The sizing helper lists every configuration that fits a given model, cheapest first. GeForce cards suit generation and rendering; the L40S and RTX PRO 6000 affordable serving; the H200 and MI355X large models; the B200 and B300 training at scale.

No. Each server holds up to 8 GPUs linked inside the server. High-speed fabrics between servers are not part of our offer today.

The SXM version, with 80 GB of HBM3 and NVLink 4 at 900 GB/s per GPU.

Root SSH access to your server. It is your machine for the month.

The ones you choose. When you order, you can have the GPU driver, CUDA or ROCm and Docker installed, or start from a bare OS; either way you manage them afterwards as root. The NVIDIA and AMD ROCm guides walk you through it.

Yes. Install Docker or another container runtime with GPU support. See the containers guide.

No managed service. When you order, you can have vLLM, SGLang or Ollama with Open WebUI installed with the model of your choice, plus ComfyUI or JupyterLab, at no extra cost. They run on your own server, which you manage as root.

You are: keep the software up to date, protect your SSH keys and expose only the services you need. The SSH guide covers the basics.

In under 10 minutes once your payment is confirmed on-chain. You receive the server address and root SSH access. How long the confirmation takes depends on the network and the fee you paid.

No. CryptGPU runs no KYC: no ID documents, no selfie and no identity verification.

On the deploy page: choose the GPU and the size, add your email (your SSH public key is optional), and pick a cryptocurrency. Your order then gets a private page with the payment details, the live payment status and the progress of your server. No account is needed.

No. You pay one month at a time and renew when you choose.

Order the new configuration and move your data; the old server runs until its term ends. Moving data between servers is your responsibility.

Yes. The pricing API returns our GPU catalog, prices and quotes as JSON, with no key needed.

If you renew, nothing changes. If you do not, the server stops at the end of the paid month and its disks are wiped. Back up your data before the term ends.

Payments are non-refundable once the server has been delivered, except where the law requires otherwise. If we cannot deliver a server you ordered, we refund the payment in crypto to an address you give us, or credit it, as you prefer.

No, unless agreed in writing beforehand. See the Acceptable Use Policy.

Yes. Advanced GPUs are export-controlled, and you must comply with export-control and sanctions laws when you use the service.

No SLA is published. Support is provided through the contact form.

This website has no analytics, no advertising trackers and no third-party scripts. It sets one functional cookie when you dismiss the announcement bar. Details in the privacy policy.