Metadata-Version: 2.4
Name: beaver-ai
Version: 0.1.0
Summary: Python SDK for the Beaver AI Gateway
Requires-Python: >=3.8
Description-Content-Type: text/markdown
Requires-Dist: requests>=2.25.0

# Beaver AI Python SDK

Python SDK for the Beaver AI Gateway.

## Installation

```bash
pip install beaver-ai
```

## Quick Start

```python
from beaver_ai import Beaver

beaver = Beaver(api_key="your-api-key")

response = beaver.chat.completions.create(
    model="gemini-2.5-flash",
    messages=[
        {"role": "user", "content": "Hello, how are you?"}
    ]
)

print(response["choices"][0]["message"]["content"])
```

## Configuration

### Constructor Options

```python
beaver = Beaver(
    api_key="your-api-key",      # Required
    base_url="http://localhost:8080"  # Optional, defaults to http://localhost:8080
)
```

## API Reference

### Chat Completions

Create a chat completion request.

```python
response = beaver.chat.completions.create(
    model="gemini-2.5-flash",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "What is the capital of France?"}
    ],
    temperature=0.7,           # Optional
    max_output_tokens=1000,    # Optional
    top_p=1.0                  # Optional
)
```

#### Request Parameters

- `model` (str, required): Model identifier
- `messages` (list[dict], required): List of message dictionaries with `role` and `content`
- `temperature` (float, optional): Sampling temperature (0-2)
- `max_output_tokens` (int, optional): Maximum tokens to generate
- `top_p` (float, optional): Nucleus sampling parameter

#### Response Format

```python
{
    "id": str,
    "model": str,
    "choices": [
        {
            "index": int,
            "message": {
                "role": "assistant",
                "content": str
            },
            "finish_reason": str | None
        }
    ],
    "usage": {  # Optional
        "prompt_tokens": int,
        "completion_tokens": int,
        "total_tokens": int
    }
}
```

## Error Handling

The SDK raises `BeaverError` for non-2xx responses.

```python
from beaver_ai import Beaver, BeaverError

beaver = Beaver(api_key="your-api-key")

try:
    response = beaver.chat.completions.create(
        model="gemini-2.5-flash",
        messages=[{"role": "user", "content": "Hello"}]
    )
except BeaverError as e:
    print(f"Error {e.status}: {e.message}")
    print(f"Code: {e.code}")
```

### Error Properties

- `status` (int): HTTP status code
- `code` (str | None): Beaver error code
- `message` (str): Error message

## Custom Base URL

```python
beaver = Beaver(
    api_key="your-api-key",
    base_url="https://api.beaver.example.com"
)
```

## Model Support

The SDK accepts model identifiers as strings. Currently supported and documented model: `gemini-2.5-flash`.

## Type Hints

The SDK includes type hints for better IDE support.

```python
from beaver_ai import ChatCompletionRequest, ChatCompletionResponse
```

## License

MIT
