README

LLM API Basic Scripts

This repository demonstrates the basics of working with multiple Large Language Model (LLM) providers. It shows how to send requests to different models in both streaming and non-streaming modes.

Supported providers

• Anthropic (Claude)
• Azure OpenAI
• Google Gemini
• Mistral
• OpenAI

---

Structure

• anthropic/ → examples with Claude models
• azure_open_ai/ → examples with Azure-hosted OpenAI models
• google/ → examples with Gemini models
• mistral/ → examples with Mistral models
• open_ai/ → examples with OpenAI models
• dev_chat/ → interactive developer chat with provider selection and transcripts
• .env.example → template for environment variables

Each provider directory contains scripts for:

• Streaming responses (real-time output)
• Non-streaming responses (full output returned at once)

The providers/ folder includes lightweight provider classes for each API (Anthropic, OpenAI, Azure, Mistral, Google Gemini). They unify the usage, so you can plug them into the dev_chat interface.

---

Setup

1. Clone the repository.

2. Create a Python virtual environment:

   python -m venv .venv

3. Activate the virtual environment:

   • On Linux/macOS:

     source .venv/bin/activate

   • On Windows (PowerShell):

     .venv\Scripts\Activate.ps1

4. Install dependencies:

   pip install -r requirements.txt

5. Copy .env.example to .env and fill in your API keys.

---

Dev Chat

The dev_chat tool allows you to interactively chat with any supported provider. Features include:

• Provider selection at runtime
• Support for conversation history
• 10-turn session limit
• Markdown transcript export

Run the chat:

python dev_chat/dev_chat.py

or (if installed globally):

py dev_chat/dev_chat.py

Example screenshot

---

requirements

anthropic
openai
mistralai
google-genai
python-dotenv
rich

Credits

Examples inspired by and adapted from official SDK documentation of:

• OpenAI
• Azure OpenAI
• Anthropic
• Google Generative AI
• Mistral