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

# Overview

> Pay-per-request AI inference over x402

The x402 Inference API exposes Dolphin AI chat completions behind an x402 paywall. There are no API keys and no subscriptions — each chat request is paid for with a single USDC micropayment, authorized by a wallet signature.

## Base URL

The inference API is served by the Meridian facilitator:

```
https://api.mrdn.finance
```

## Endpoints

<CardGroup cols={2}>
  <Card title="Chat Completion" icon="message" href="/api-reference/inference/chat">
    `POST /v1/inference/chat` — paid, streams SSE
  </Card>

  <Card title="List Models" icon="microchip" href="/api-reference/inference/models">
    `GET /v1/inference/models` — free
  </Card>

  <Card title="List Templates" icon="list" href="/api-reference/inference/templates">
    `GET /v1/inference/templates` — free
  </Card>
</CardGroup>

## How payment works

`POST /v1/inference/chat` is protected by the standard x402 flow:

1. Call the endpoint with no payment — you receive `402 Payment Required` with the accepted payment options (price, network, asset, receiver).
2. Sign an EIP-3009 `TransferWithAuthorization` for the quoted amount (a gasless signature — no transaction, no gas needed).
3. Retry the same request with the signed payment base64-encoded in the `X-PAYMENT` header.
4. The server verifies and settles the payment through the Meridian facilitator, then streams the completion.

Any x402 client handles this automatically — see [Programmatic API Access](/api-reference/programmatic-access) for `x402-fetch` / `x402-axios` examples that work against this API unchanged.

## Models

| ID                  | Name            |
| ------------------- | --------------- |
| `dolphinserver:24B` | Dolphin 24B     |
| `dolphinserver2:6b` | Dolphin X1 Nano |

## Templates

| Name            | Label         |
| --------------- | ------------- |
| `logical`       | Logical       |
| `creative`      | Creative      |
| `summary`       | Summarize     |
| `code-beginner` | Code Beginner |
| `code-advanced` | Code Advanced |

<Note>
  Models and templates are sourced live from the upstream provider — always
  fetch the current lists from `GET /v1/inference/models` and `GET /v1/inference/templates`
  rather than hardcoding them.
</Note>