Quick start
From sign-up to a running GPU in four steps
Per-second billing from a prepaid balance. No subscription, no contract.
- 1
Create an account
Google or email. Verify your email, then add an SSH key in the console so your machines are reachable.
Create an account - 2
Top up your balance
Card or crypto, any amount from $5. First top-up of $10 or more: we add $3.80, your first hour of RTX PRO 6000 on us. One hour is held when you launch and the unused part comes back to the second when you stop.
Billing → - 3
Launch what you need
Confidential VM
A whole VM in an Intel TDX trust domain with an NVIDIA GPU in confidential mode. Root over SSH, attestation you generate yourself.
Deploy a confidential VM →Standard GPU
No enclave, for data that isn't sensitive. Templates or your own Docker image, from the console, the API or the CLI.
Browse standard GPUs →TEE inference API
OpenAI-compatible API to models running in Intel TDX enclaves. Change the base URL, keep your SDK. Requests pass through our relay, which sees them in clear (TLS ends at our host). To keep plaintext off our servers, run your own inference server in a Confidential VM and terminate TLS inside its trust domain.
See the models →Inference in Python
from openai import OpenAI client = OpenAI(base_url="https://api.voltagegpu.com/v1", api_key="YOUR_API_KEY") r = client.chat.completions.create( model="google/gemma-4-31B-turbo-TEE", messages=[{"role": "user", "content": "Hello"}], ) print(r.choices[0].message.content)Standard GPU from the terminal
pip install voltagegpu-cli voltagegpu machines --standard voltagegpu deploy standard --machine <machine_id> voltagegpu ssh <pod_id> - 4
Check what you got
On a confidential VM, generate the Intel TDX and NVIDIA attestation yourself and verify it offline with voltage-verify. On a standard GPU, one read-only command compares the machine with its listing.
curl -fsSL https://voltagegpu.com/hwcheck.sh | bash -s -- --gpu "RTX 4090" --gpus 1How to verify the attestation → How the hardware check works →