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

# Route Claude Requests Through Flatkey via Anthropic SDK

> Route Anthropic Claude requests through Flatkey by setting the base URL in the Anthropic SDK. Access Claude at up to 50% off official pricing.

You can access Claude models at up to 50% off official Anthropic pricing by routing the Anthropic Python SDK through Flatkey. Set the `base_url` parameter once and your existing SDK code works as-is — no other changes required.

## Installation

```bash pip theme={null}
pip install anthropic
```

## Basic setup

```python python theme={null}
import os
import anthropic

client = anthropic.Anthropic(
    api_key=os.environ["FLATKEY_API_KEY"],
    base_url="https://router.flatkey.ai",
)
```

<Note>
  When using the Anthropic SDK, set `base_url` to `https://router.flatkey.ai` (without `/v1`). The SDK appends the path automatically.
</Note>

## Sending a message

```python python theme={null}
message = client.messages.create(
    model="claude-sonnet-4-5",
    max_tokens=1024,
    messages=[
        {"role": "user", "content": "Write a Python function to reverse a linked list."}
    ],
)

print(message.content[0].text)
```

## System prompts

```python python theme={null}
message = client.messages.create(
    model="claude-opus-4-5",
    max_tokens=2048,
    system="You are a senior software engineer. Provide concise, idiomatic code.",
    messages=[
        {"role": "user", "content": "Implement a rate limiter in Python."}
    ],
)

print(message.content[0].text)
```

## Streaming

```python python theme={null}
with client.messages.stream(
    model="claude-haiku-4-5-20251001",
    max_tokens=512,
    messages=[{"role": "user", "content": "Explain how HTTPS works."}],
) as stream:
    for text in stream.text_stream:
        print(text, end="", flush=True)
print()
```

## Multi-turn conversations

```python python theme={null}
conversation = [
    {"role": "user", "content": "What is the capital of France?"},
    {"role": "assistant", "content": "The capital of France is Paris."},
    {"role": "user", "content": "What is its population?"},
]

response = client.messages.create(
    model="claude-sonnet-4-5",
    max_tokens=256,
    messages=conversation,
)

print(response.content[0].text)
```

## Available Claude models

| Model                       | Best for                             | Input price (after bonus) |
| --------------------------- | ------------------------------------ | ------------------------- |
| `claude-opus-4-5`           | Complex reasoning, long-form writing | \~\$3 / 1M tokens         |
| `claude-sonnet-4-5`         | Balanced quality and speed           | \~\$1.6 / 1M tokens       |
| `claude-haiku-4-5-20251001` | Fast, cost-efficient tasks           | \~\$0.53 / 1M tokens      |

See the [Model Directory](/reference/model-list) for all Claude models and current pricing.

<Tip>
  You can also use Claude models with the OpenAI SDK through Flatkey — just set `model="claude-sonnet-4-5"` in a `client.chat.completions.create()` call. Use whichever SDK fits your project.
</Tip>