Enterprise AI,
Without the Compromise.
Secure, private, and localized AI chat and API infrastructure. Backed by client-side ECIES encryption and managed compute.
Intelligence Tailored to the Task
Opaque, highly-optimized routing ensures you get the right reasoning model for the right price, instantly.
Fast
Optimized for low-latency tasks, drafting, and instant summarization. Uncapped usage for smooth workflows.
Pro
High-fidelity reasoning and complex instruction following for heavy enterprise workloads.
Deep Research
Multi-step autonomous reasoning with built-in prolog loops for deep analytical tasks.
Cryptographic Privacy You Can Verify
We engineered our infrastructure so that neither DCP nor our cloud provider can inspect your conversation context.
Passkey-Derived Sessions
Every chat session generates a dedicated AES key anchored directly to your hardware authenticator using WebAuthn.
ECIES Envelope Wrapping
Keys are transmitted inside an ECIES envelope decoded strictly in memory, preventing raw keys from ever surfacing in GCP request logs.
Zero-Knowledge Context
Context history is stored in ciphertext. We maintain the database, but only client-authenticated sessions hold the decryption keys.
Client-to-GPU TEE
Full end-to-end encryption extending from your browser directly into hardware-isolated Confidential GPU enclaves.
// 1. Session AES key derived via WebAuthn PRF
const sessionKey = await crypto.subtle.generateKey(
{ name: "AES-GCM", length: 256 },
true,
["encrypt", "decrypt"]
);
// 2. Wrap AES key in an ECIES envelope for runtime inference
const encryptedKeyEnvelope = await eciesEncrypt(
sessionKeyRaw,
backendPublicVerificationKey
);
// 3. Dispatch payload - raw AES keys never enter cloud provider logs or databases
await fetch("/api/v1/chat/completions", {
headers: { "X-DCP-Key-Envelope": encryptedKeyEnvelope },
body: encryptedContextPayload
});Transparent, Capability-Based Pricing
Choose a workspace subscription for your team, or pay-as-you-go for direct API integration.
Workspace Subscriptions
Standard
Ideal for everyday tasks and standard workloads.
75 / month
- ✓ Unlimited Fast Model usage
- ✓ Pro Model (Standard 4-hour quota)
- ✓ Graceful shift to Fast when quota is met
Developer
For heavy analytics, research, and data science.
100 / month
- ✓ Pro Model (High 4-hour quota)
- ✓ Deep Research Model access
- ✓ Managed Jupyter Compute included
Enterprise & Compute
Ideal for everyday tasks and standard workloads.
500 / month
- ✓ Pro Model (High 4-hour quota)
- ✓ Deep Research Model access
- ✓ Managed Jupyter Compute included
API Pay-As-You-Go
150 / 1M TokensAPI usage is billed per token. Cost is calculated dynamically based on the capability weight of the chosen endpoint.
Take Zero-Knowledge AI on the Go
Our native mobile apps will bring full end-to-end encryption and seamless workspace syncing to your pocket.