Examples

Basic Chat Completion

from openwebui_client import OpenWebUIClient

client = OpenWebUIClient(
    api_key="your-openwebui-api-key",
    base_url="http://localhost:5000"
)

response = client.chat.completions.create(
    model="gpt-4",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "What can you tell me about OpenWebUI?"}
    ]
)
print(response.choices[0].message.content)

Chat Completion with Files

from openwebui_client import OpenWebUIClient

client = OpenWebUIClient(
    api_key="your-openwebui-api-key",
    base_url="http://localhost:5000"
)

# Read file content
with open("document.pdf", "rb") as f:
    file_content = f.read()

# Create chat completion with file attachment
response = client.chat.completions.create(
    model="gpt-4",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Summarize the attached document."}
    ],
    files=[file_content]
)

Using Function Calling / Tools

Direct Tool Usage

from openwebui_client import OpenWebUIClient

client = OpenWebUIClient(
    api_key="your-openwebui-api-key",
    base_url="http://localhost:5000"
)

# Define a tool schema directly
response = client.chat.completions.create(
    model="your_model",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "What is the current time?"}
    ],
    tools=[
        {
            "type": "function",
            "function": {
                "name": "get_current_time",
                "description": "Get the current time.",
                "parameters": {
                    "type": "object",
                    "properties": {},
                    "required": [],
                },
            },
        }
    ],
)

# Check if the model used tools
if response.choices[0].message.tool_calls:
    tool_call = response.choices[0].message.tool_calls[0]
    print(f"Tool called: {tool_call.function.name}")

Using the Tool Registry

from openwebui_client import OpenWebUIClient
from pathlib import Path

client = OpenWebUIClient(
    api_key="your-openwebui-api-key",
    base_url="http://localhost:5000"
)

# Define tool functions
def get_weather(location: str, unit: str = "celsius") -> str:
    """Get the current weather in a given location.

    Args:
        location: The location to get weather for
        unit: The temperature unit to use (celsius or fahrenheit)

    Returns:
        str: A string describing the current weather
    """
    return f"The weather in {location} is sunny and 25°{unit[0]}"

def get_forecast(location: str, days: int = 1) -> str:
    """Get the weather forecast for a location.

    Args:
        location: The location to get the forecast for
        days: Number of days to forecast

    Returns:
        str: A string describing the forecast
    """
    return f"The forecast for {location} for the next {days} day(s) is sunny."

# Register tools with the client
client.tool_registry.register(get_weather)
client.tool_registry.register(get_forecast)

# Use chat_with_tools for automatic tool handling
response = client.chat_with_tools(
    messages=[
        {"role": "system", "content": "You are a helpful weather assistant."},
        {"role": "user", "content": "What's the weather like in Toronto?"}
    ],
    max_tool_calls=5,
)

print(response)  # Will contain the final response after any tool calls

# You can also include files with tool calls
file_path = Path("weather_data.txt")
response_with_file = client.chat_with_tools(
    messages=[
        {"role": "system", "content": "You are a helpful weather assistant."},
        {"role": "user", "content": "Analyze the weather data in the attached file."}
    ],
    files=[file_path],
    max_tool_calls=5,
)
print(response.choices[0].message.content)

File Upload and Management

from openwebui_client import OpenWebUIClient

client = OpenWebUIClient(
    api_key="your-openwebui-api-key",
    base_url="http://localhost:5000"
)

# Read file content
with open("document.pdf", "rb") as f:
    file_content = f.read()

# Upload a single file
file_obj = client.files.create(
    file=file_content,
    file_metadata={"purpose": "assistants"}
)
print(f"File uploaded with ID: {file_obj.id}")

# Upload multiple files
with open("another_doc.pdf", "rb") as f2:
    file_content2 = f2.read()

file_objects = client.files.create(
    files=[(file_content, {"purpose": "assistants"}),
           (file_content2, {"purpose": "assistants"})]
)
for i, file_obj in enumerate(file_objects):
    print(f"File {i+1} uploaded with ID: {file_obj.id}")