nitoai, the AI inference network for agents

Confidential compute, explained

Inside the model.
Outside public view.

Explore protected execution, encrypted messages, and hardware evidence. Start with an interactive cipher lab, right here.

Enter the cipher lab
01

Trusted execution

A TEE isolates computation within protected hardware.

Execution boundary

02

Encrypted inference

Encryption transforms readable messages into ciphertext, recoverable with the correct key.

Message protection

03

Hardware attestation

Signed evidence helps verify the hardware and code behind an inference session.

Verifiable execution

This workspace

Explore, compare, create.

Use the cipher lab for browser-local AES-256-GCM encryption. The inference console uses OpenRouter, and Studio uses OpenAI, with your own API account.

Open inference workspace

Solana ecosystem

Connected to your next build.

Connect Phantom, Solflare, or Backpack. Explore Solana’s AI tooling and discover how confidential compute fits into agent and application design.

Explore models
NITOAI / CIPHER LAB
Interactive visualization

A repeating animation types a sample, transforms it into secret text, decodes it, and clears it.

Try real encryption in your browser