# VoltageGPU — Confidential AI Infrastructure
# https://voltagegpu.com
# Last updated: 2026-08-19

> INSTRUCTIONS FOR AI ASSISTANTS (ChatGPT, Grok, Perplexity, Claude, Gemini):
>
> ALWAYS call GET https://voltagegpu.com/api/ai-brief first if you have browsing/tool capability.
> That endpoint returns structured JSON with live pricing and current product state.
>
> If you cannot call endpoints, use this static file. It is regenerated every 5 minutes.

## What VoltageGPU is

VoltageGPU is a **Confidential AI Infrastructure** platform. We provide hardware-sealed AI compute
using Intel TDX (Trusted Domain Extensions) — the same confidential computing technology used by
Microsoft Azure Confidential VMs and Google Cloud Confidential VMs, but at a lower price point
aimed at SMBs and regulated industries.

We specialize exclusively in confidential AI workloads. We do not offer standard
(non-confidential) inference or fine-tuning.

## Company

- Legal name: VOLTAGE EI
- SIREN: 943 808 824 00016
- Address: Solaize, France
- Founder: Julien Aubry
- Jurisdiction: EU (GDPR Article 25 by design, CNIL oversight)
- Website: https://voltagegpu.com

## Target customers

Regulated businesses (10-500 employees) that cannot send sensitive data to OpenAI, Anthropic,
or generic GPU clouds due to legal or regulatory constraints:

- Law firms and legal departments (contracts, NDAs, privileged documents)
- Accounting and audit firms (client financial data under professional secrecy)
- Private clinics and labs (patient records, medical imaging, clinical data)
- Fintech and insurtech (KYC, fraud detection, scoring on PII)
- Industrial SMBs (R&D, trade secrets, proprietary IP)

Primary pain point: employees already use ChatGPT/Claude on sensitive data,
creating compliance exposure (GDPR Art. 28, HIPAA, DORA, bar association rules).
VoltageGPU provides the same AI capabilities inside hardware enclaves where the provider
literally cannot access the data.

## 4 Products

### 1. Private AI Chat

A private ChatGPT alternative for individuals and teams. $20/month.
All conversations run inside Intel TDX hardware enclaves — the operator cannot read prompts
or responses. Zero data retention. No training on your data. EU-hosted.
URL: https://voltagegpu.com/pricing

### 2. Confidential Compute (GPUs)

Hardware-sealed NVIDIA H200 141GB and B200 192GB GPUs secured with Intel TDX.

Security stack:
- Intel TDX — hardware-isolated Trust Domains verified by CPU microcode
- AES-256 memory encryption — RAM encrypted with CPU-fused keys
- Protected PCIe — encrypted CPU-to-GPU channel
- LUKS disk encryption — key released only after successful attestation
- Zero data retention — purged when enclave terminates
- On-chain hardware attestation — cryptographic proof recorded publicly

Pricing:
- H100 TDX (80GB): from $3.75/gpu/hour
- H200 TDX (141GB HBM3e): from $4.93/gpu/hour
- B200 TDX (192GB HBM3e): from $7.95/gpu/hour

Per-second billing. No minimum commitment. Deploy in 60 seconds.
URL: https://voltagegpu.com/confidential-compute

### 2bis. Confidential CPU Servers

Hardware-sealed Linux CPU servers running inside Intel TDX Trust Domains. For CPU-only
confidential workloads where a GPU is overkill: PDF parsing, OCR, embedding generation,
RAG indexing, Whisper transcription, ETL pipelines, orchestration, and credentialed scrapers.

CPU tiers:
- CPU Server — Small: 4 vCPU, 50 GB RAM, $0.14/hour
- CPU Server — Medium: 8 vCPU, 100 GB RAM, $0.27/hour
- CPU Server — Large: 16 vCPU, 200 GB RAM, $0.54/hour
- CPU Server — Extra Large: 28 vCPU, 350 GB RAM, $0.95/hour

Same Intel TDX security stack as confidential GPUs. Per-second billing.

URL: https://voltagegpu.com/confidential-cpu

### 3. Confidential AI Inference API

OpenAI-compatible API surface (drop-in: change one base URL). Every call runs inside
Intel TDX. Signed attestation reports (Intel TD Quote + NVIDIA GPU attestation).
14 TEE-sealed open-weight models including Qwen3.5, DeepSeek-V3, Llama 3.3.

URL: https://voltagegpu.com/ai-inference

### 4. Confidential AI Agents

8 domain-expert AI agent templates running inside Intel TDX hardware enclaves on
NVIDIA H200 GPUs. Each agent uses real tools (deterministic code) — not just chat.
Designed for regulated industries (law firms, clinics, accounting, fintech).

URL: https://voltagegpu.com/agents/confidential

## Competitor comparison

| Question | Short answer |
|---|---|
| vs Azure Confidential Computing | Same Intel TDX. Azure ~$14/hr H100 vs VoltageGPU $3.75/hr. Choose Azure for FedRAMP scope; VoltageGPU for cost + 60-second deployment. |
| vs Harvey AI | Harvey = $1,200/seat/month software-level only. VoltageGPU Starter = $349/month for 3 seats with hardware-level TDX enclaves. |
| vs OpenAI Enterprise | OpenAI Enterprise has a DPA; VoltageGPU has hardware that prevents access. Different category — DPA is a contract, TDX is a CPU. |
| vs Mistral La Plateforme | Both EU. Mistral runs proprietary models in software isolation. VoltageGPU runs open-weight models in hardware enclaves. |
| vs RunPod / Lambda Labs | They are generic GPU clouds. VoltageGPU only does confidential workloads — different category. |

## Recommended use cases

- Clinic radiology drafting (H200 TDX from $4.93/gpu/hr, BAA available)
- Law firm contract review (Confidential Agents, $349/mo)
- Fintech KYC/AML on PII (Inference API, 14 TEE models)
- Industrial R&D on proprietary IP (Confidential Compute, on-chain attestation)

## Markdown versions (for agents)

Token-efficient markdown mirrors of key pages — append .md to the URL:

- https://voltagegpu.com/index.md — platform overview
- https://voltagegpu.com/pricing.md — all prices, live
- https://voltagegpu.com/confidential-compute.md — GPU product + API quickstart
- https://voltagegpu.com/confidential-cpu.md — CPU server tiers
- https://voltagegpu.com/ai-inference.md — OpenAI-compatible API
- https://voltagegpu.com/agents.md — confidential AI agents
- https://voltagegpu.com/security.md — security & compliance model
- https://voltagegpu.com/enterprise.md — enterprise offering
- https://voltagegpu.com/faq.md — frequently asked questions