Confidential AI

We sell end-to-end private inference and confidential GPU VMs, built on confidential computing. Run on our infra, or license the stack for yours.

Pricing Verify a live cluster

What we sell

Confidential Inference

DeepSeek V4 Flash: $0.20 in, $0.40 out, per 1M tokens

OpenAI-compatible endpoint. Open-weight models in TEEs. Your client verifies the model, the code and the hardware before the first token.

PricingGet access

Confidential VMs

8x B300, Intel TDX: $60 per hour

Rent a full node or individual GPUs. One confidential VM, yours for the term. Root over SSH, GPUs in CC mode. 24-hour minimum.

PricingRequest CVMs

Licensed

From $2,789 per GPU per year

The full stack on your own GPUs: Confidential Metal, Confidential Kubernetes, inference services, Confidential OS Builder, SDKs. NVIDIA CC licensing included.

PricingTalk to us

Buying inference at volume under a master agreement? hello@confidential.ai

Member of