llm-client

A lightweight client wrapper for OpenAI-compatible chat APIs with streaming, tool calls, and simple cost tracking.

Setup

1. Copy src/llm_client/clients/openai_client.py into your project
2. Create a configuration file based on src/llm_client/conf/llm.example.yaml
3. Update the config with your API endpoint and key

Usage

from llm_client import OpenAIClient

client = OpenAIClient()
messages = [{"role": "user", "content": "Say hi!"}]
print(client.send_messages(messages, model="deepseek"))