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

# Introduction

> A confidential AI inference API you can verify. Every response is served from a TEE and backed by an attestation report and a signed receipt.

<img className="block" src="https://mintcdn.com/phala-47f19b07/YXikfPBwb_GQq6hw/images/banner.png?fit=max&auto=format&n=YXikfPBwb_GQq6hw&q=85&s=07f49e8efa87f23021202245e20cd068" alt="Confidential AI" width="1500" height="1000" data-path="images/banner.png" />

This is a confidential AI inference API. Every response is served by a gateway that runs inside a
Trusted Execution Environment (TEE), and comes with cryptographic evidence you can verify: an
attestation report proving which code served you, and a signed receipt binding your request and
response to that attested workload.

## Why it is different

<CardGroup cols={3}>
  <Card title="Verifiable privacy" icon="shield-check">
    The gateway runs in a TEE and proves it with a hardware-backed
    [attestation report](/confidential-ai/attestation-report). It does not retain request or response
    bodies.
  </Card>

  <Card title="Confidential models" icon="lock">
    Confidential models run on [verified TEE providers](/confidential-ai/providers), with the channel
    bound before your prompt is forwarded.
  </Card>

  <Card title="Drop-in compatible" icon="plug">
    Point your OpenAI or Anthropic SDK at the base URL. Same request and response shapes you already
    use.
  </Card>
</CardGroup>

## Pick your path

<CardGroup cols={2}>
  <Card title="Build with the API" icon="code" href="/get-started/quickstart">
    For developers. Call confidential and standard models through one endpoint, with attestation and
    signed receipts on every response.
  </Card>

  <Card title="Use the chat app" icon="message" href="/chat/overview">
    For people who want a private AI assistant. End-to-end encrypted chat with confidential models and
    private document search.
  </Card>
</CardGroup>

## Verify, don't trust

You do not have to take the API's word. After one request you can check, locally, that:

1. The gateway is a specific reviewed workload running in a genuine TEE.
   [Attestation report →](/confidential-ai/attestation-report)
2. The response came from that workload, was not altered, and (for a confidential model) ran on a
   verified upstream enclave. [Receipts →](/confidential-ai/receipts)

See [How confidential AI works](/confidential-ai/how-it-works) for the model, or
[Verify a response](/guides/verify-a-response) to run the checks.

## Start building

<Steps>
  <Step title="Get an API key">
    Create a key in the [dashboard](https://www.redpill.ai/dashboard) and store it as an environment
    variable. See [Authentication](/get-started/authentication).
  </Step>

  <Step title="Make your first call">
    Point an OpenAI client library at `https://api.redpill.ai/v1`. See the
    [Quickstart](/get-started/quickstart).
  </Step>

  <Step title="Verify the response">
    Fetch the attestation report and the receipt, then check the signatures and hashes.
    [Verify a response](/guides/verify-a-response).
  </Step>
</Steps>

<CardGroup cols={3}>
  <Card title="Quickstart" icon="bolt" href="/get-started/quickstart" />

  <Card title="Models" icon="layer-group" href="/get-started/models" />

  <Card title="API Reference" icon="book" href="/api-reference/overview" />
</CardGroup>