zenllm 0.1.4

A zen, simple, and unified API to prompt LLMs from Anthropic, Google, OpenAI, and more, using only the requests library.

🧘‍♂️ ZenLLM

The zen, simple, and unified API to prompt LLMs from different providers using a single function call.

Our Goal: No fancy dependencies, no SDK bloat. Just requests and your API keys. Easily switch between models and providers without changing your code.

✨ Key Features

• Unified API: A single prompt() function for Anthropic, Google, and OpenAI.
• Easy Model Switching: Change models and providers with a single model argument.
• Lightweight: Built with only the requests library. No heavy SDKs to install.
• Simple Authentication: Just set your provider's API key as an environment variable.

🚀 Installation

pip install zenllm

💡 Usage

First, make sure you've set your API key as an environment variable (e.g., export OPENAI_API_KEY="your-key").

Basic Prompt

The default model is gpt-4.1. You can change this by setting the ZENLLM_DEFAULT_MODEL environment variable.

from zenllm import prompt

# Uses the default model (either the fallback or ZENLLM_DEFAULT_MODEL)
response = prompt("Why is the sky blue?")
print(response)

Using another Provider (Google's Gemini)

Simply change the model name to use a different provider.

from zenllm import prompt

response = prompt(
    "Why is the sky blue?",
    model="gemini-2.5-pro",
    system_prompt="Reply only in French."
)
print(response)

Using Anthropic

from zenllm import prompt

response = prompt(
    "Tell me a three sentence bedtime story about a unicorn.",
    model="claude-sonnet-4-20250514"
)
print(response)

Streaming Responses

Set stream=True to get a generator of response chunks.

from zenllm import prompt

response_stream = prompt(
    "Tell me a story about a robot.", 
    model="claude-sonnet-4-20250514", 
    stream=True
)
for chunk in response_stream:
    print(chunk, end="", flush=True)

✅ Supported Providers

Provider	Environment Variable	Model Prefix	Example Models
Anthropic	ANTHROPIC_API_KEY	claude	claude-sonnet-4-20250514, claude-opus-4-20250514
Deepseek	DEEPSEEK_API_KEY	deepseek	deepseek-chat, deepseek-reasoner
Google	GEMINI_API_KEY	gemini	gemini-2.5-pro, gemini-2.5-flash
OpenAI	OPENAI_API_KEY	gpt	gpt-4.1

Note: Streaming is not yet supported for OpenAI with this library.

📜 License

MIT License - Copyright (c) 2025 Koen van Eijk