LLM Handler

A unified Python package for handling LLM API calls across multiple providers (OpenAI, Anthropic, and Google).

Features

• Unified interface for multiple LLM providers
• Async and sync support
• Token limit validation
• Consistent token counting across providers
• Built-in error handling
• Environment variable configuration

Installation

pip install llm-handler

Usage

Async Usage

import asyncio
from llm_handler import LLMHandler

async def main():
    handler = LLMHandler()
    
    response = await handler.get_completion(
        prompt="Tell me a joke",
        model="gpt-4o",
        temperature=0.7,
        max_tokens=100
    )
    
    print(f"Response: {response.text}")
    print(f"Prompt tokens: {response.prompt_tokens}")
    print(f"Completion tokens: {response.completion_tokens}")

asyncio.run(main())

Synchronous Usage

from llm_handler import LLMHandler

handler = LLMHandler()
response = handler.get_completion_sync(
    prompt="Tell me a joke",
    model="gpt-4o",
    temperature=0.7,
)
print(response.text)

Supported Models

OpenAI

• gpt-4o
• gpt-4o-mini
• o1-preview
• o1-mini

Anthropic

• claude-3-opus-latest
• claude-3-sonnet-latest

Google

• gemini-1.5-pro

Configuration

Create a .env file in your project root:

OPENAI_API_KEY=your_openai_key
ANTHROPIC_API_KEY=your_anthropic_key
GOOGLE_API_KEY=your_google_key

License

MIT License - see LICENSE file for details