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

# Anthropic Format Overview

> Use Claude models with the official Anthropic SDK — point it at Samurai AI and your existing code works unchanged.

## What is Anthropic Native Format?

Samurai AI supports **two API formats** side-by-side:

| Format               | Endpoint                    | Auth Header                    | SDK             |
| -------------------- | --------------------------- | ------------------------------ | --------------- |
| OpenAI-compatible    | `POST /v1/chat/completions` | `Authorization: Bearer sk-...` | `openai`        |
| **Anthropic-native** | **`POST /v1/messages`**     | **`x-api-key: sk-...`**        | **`anthropic`** |

If you already use the official `anthropic` Python or TypeScript SDK — or call `api.anthropic.com` directly — you can switch to Samurai AI **by changing one line**: the `base_url`.

```python theme={null}
# Before (Anthropic Cloud)
from anthropic import Anthropic
client = Anthropic(api_key="sk-ant-...")

# After (Samurai AI — everything else stays the same)
from anthropic import Anthropic
client = Anthropic(
    base_url="https://www.samuraiapi.in/api",
    api_key="YOUR_SAMURAI_KEY"
)
```

<Info>
  **Base URL for Anthropic format**: `https://www.samuraiapi.in/api`

  The Anthropic SDK appends `/v1/messages` automatically. Do not include it in the base URL.
</Info>

***

## Supported Claude Models

| Model ID                     | Context | Best For                                   |
| ---------------------------- | ------- | ------------------------------------------ |
| `claude-opus-4-5`            | 200K    | Complex analysis, research, long documents |
| `claude-sonnet-4-5`          | 200K    | Balanced performance, most tasks           |
| `claude-3-5-sonnet-20241022` | 200K    | Coding, reasoning, instruction-following   |
| `claude-3-5-haiku-20241022`  | 200K    | Fast, low-latency responses                |
| `claude-3-opus-20240229`     | 200K    | Deep reasoning, nuanced writing            |

<Tip>
  You can also use Claude models through the OpenAI-compatible `/v1/chat/completions` endpoint — Samurai AI converts the format automatically. Use whichever endpoint matches your existing SDK.
</Tip>

***

## Key Differences from OpenAI Format

### 1. System prompt is a top-level field

```python theme={null}
# OpenAI format — system message inside messages array
messages = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Hello!"}
]

# Anthropic format — system is a separate top-level parameter
response = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1024,
    system="You are a helpful assistant.",   # <-- top-level
    messages=[
        {"role": "user", "content": "Hello!"}
    ]
)
```

### 2. `max_tokens` is required

The Anthropic API always requires `max_tokens`. There is no default.

### 3. Response is a `Message` object, not a `ChatCompletion`

```json theme={null}
// OpenAI response
{
  "object": "chat.completion",
  "choices": [{ "message": { "role": "assistant", "content": "..." } }]
}

// Anthropic response
{
  "type": "message",
  "role": "assistant",
  "content": [{ "type": "text", "text": "..." }]
}
```

### 4. Tool definitions use `input_schema` instead of `parameters`

```python theme={null}
# OpenAI format
tools = [{"type": "function", "function": {"name": "get_weather", "parameters": {...}}}]

# Anthropic format
tools = [{"name": "get_weather", "input_schema": {...}}]
```

### 5. Auth header is `x-api-key`

```bash theme={null}
# OpenAI
-H "Authorization: Bearer sk-samurai-..."

# Anthropic
-H "x-api-key: sk-samurai-..."
-H "anthropic-version: 2023-06-01"
```

***

## Quick Start

<CodeGroup>
  ```python Python theme={null}
  from anthropic import Anthropic

  client = Anthropic(
      base_url="https://www.samuraiapi.in/api",
      api_key="YOUR_SAMURAI_KEY"
  )

  message = client.messages.create(
      model="claude-3-5-sonnet-20241022",
      max_tokens=1024,
      system="You are a helpful assistant.",
      messages=[
          {"role": "user", "content": "Explain the theory of relativity in simple terms."}
      ]
  )

  print(message.content[0].text)
  print(f"Used {message.usage.input_tokens} input + {message.usage.output_tokens} output tokens")
  ```

  ```typescript TypeScript theme={null}
  import Anthropic from "@anthropic-ai/sdk";

  const client = new Anthropic({
    baseURL: "https://www.samuraiapi.in/api",
    apiKey: process.env.SAMURAI_API_KEY!,
  });

  const message = await client.messages.create({
    model: "claude-3-5-sonnet-20241022",
    max_tokens: 1024,
    system: "You are a helpful assistant.",
    messages: [
      { role: "user", content: "Explain the theory of relativity in simple terms." }
    ],
  });

  console.log(message.content[0].type === "text" ? message.content[0].text : "");
  ```

  ```bash cURL theme={null}
  curl https://www.samuraiapi.in/api/v1/messages \
    -H "x-api-key: YOUR_SAMURAI_KEY" \
    -H "anthropic-version: 2023-06-01" \
    -H "content-type: application/json" \
    -d '{
      "model": "claude-3-5-sonnet-20241022",
      "max_tokens": 1024,
      "system": "You are a helpful assistant.",
      "messages": [
        {"role": "user", "content": "Explain the theory of relativity in simple terms."}
      ]
    }'
  ```
</CodeGroup>

***

## SDK Installation

<CodeGroup>
  ```bash Python theme={null}
  pip install anthropic
  ```

  ```bash TypeScript / Node.js theme={null}
  npm install @anthropic-ai/sdk
  ```
</CodeGroup>

***

## Authentication

Use your Samurai AI API key in the `x-api-key` header. The same key works for both OpenAI-compatible and Anthropic-native endpoints.

```bash theme={null}
# Get your key from: https://www.samuraiapi.in/dashboard
x-api-key: sk-samurai-YOUR_KEY_HERE
```

The `anthropic-version: 2023-06-01` header is required by the official SDK — Samurai AI accepts it and forwards it as needed.

***

## Rate Limits & Quotas

Anthropic-format requests use the **same quotas and rate limits** as OpenAI-format requests. Your plan limits apply across both endpoints.

| Plan       | Requests/min | Notes                       |
| ---------- | ------------ | --------------------------- |
| Free       | 10 RPM       | Shared across all endpoints |
| Starter    | 60 RPM       | —                           |
| Pro        | 200 RPM      | —                           |
| Enterprise | Custom       | Contact us                  |

***

## Next Steps

<CardGroup cols={2}>
  <Card title="Messages API" icon="message" href="/api/anthropic/messages">
    Full reference for the `/v1/messages` endpoint — parameters, streaming, tool use, vision.
  </Card>

  <Card title="Function Calling" icon="wrench" href="/api/function-calling">
    Tool use with the OpenAI-compatible format (also works with Claude models).
  </Card>
</CardGroup>