> ## Documentation Index
> Fetch the complete documentation index at: https://docs.blindference.xyz/llms.txt
> Use this file to discover all available pages before exploring further.

# Introduction

> Private, verifiable, and economically accountable AI execution on Web3

# Welcome to Blindference

Blindference is a confidential AI execution layer for Web3. It enables users to submit sensitive prompts and data to frontier AI models — Groq Llama 3, Google Gemini — without ever revealing them to any single party.

## What It Does

* **Input Privacy**: User prompts are AES-256 encrypted in the browser before leaving the device
* **Access Control**: Encryption keys are split into FHE-encrypted halves and stored on-chain via [CoFHE](https://www.fhenix.io/), only decryptable by assigned quorum nodes
* **Execution Integrity**: A `1 leader + 2 verifier` quorum runs identical inference and cross-validates results
* **Economic Accountability**: Disputed results trigger on-chain verification with automatic USDC payouts via [Reineira](https://reineira.xyz/)
* **Output Privacy**: Only the user can decrypt the final result using their wallet

## Who Is This For?

<CardGroup cols={2}>
  <Card title="Compute Providers" icon="server" href="/compute/introduction">
    Run a Blindference node to earn fees executing encrypted inference jobs. GPU optional.
  </Card>

  <Card title="Agent Developers" icon="code" href="https://github.com/baync180705/blindference">
    Build agents that use Blindference for confidential inference, risk scoring, or custom AI workflows.
  </Card>
</CardGroup>

## Supported Modes

* **Confidential Text Inference**: Submit natural language prompts through encrypted channels
* **Confidential Risk Scoring**: Submit structured financial features for privacy-preserving evaluation

## System Architecture

```mermaid theme={null}
flowchart LR
    U[User Wallet] --> FE[Frontend]
    FE -->|AES encrypt prompt| ENC[(Encrypted Prompt Blob)]
    FE -->|Store prompt key| PKS[PromptKeyStore]
    FE -->|Submit request| ICL[ICL]

    ICL --> L[Leader Node]
    ICL --> V1[Verifier 1]
    ICL --> V2[Verifier 2]

    PKS --> L
    PKS --> V1
    PKS --> V2

    L --> LLM1[Groq / Gemini]
    V1 --> LLM2[Groq / Gemini]
    V2 --> LLM3[Groq / Gemini]

    L -->|Store output key for user| PKS
    L -->|Leader result| ICL
    V1 -->|Verdict| ICL
    V2 -->|Verdict| ICL

    ICL --> FE
    PKS --> FE
```

## Quick Links

<CardGroup cols={3}>
  <Card title="Run a Node" icon="server" href="/compute/quickstart">
    Start earning in 5 minutes
  </Card>

  <Card title="Build an Agent" icon="code" href="https://github.com/baync180705/blindference">
    Integrate confidential inference
  </Card>

  <Card title="Contract Addresses" icon="file-contract" href="/resources/contract-addresses">
    Arbitrum Sepolia deployments
  </Card>
</CardGroup>