Introduction

The FreeFlow package acts as an intelligent intermediary layer between your code and various AI providers. You simply provide multiple free API keys for your preferred services. FreeFlow intelligently handles rate limiting by first rotating through available keys within the same provider. If a provider is fully exhausted, it seamlessly switches to the next available provider. This strategy maximizes your continuous uptime and ensures you get the most out of every free tier without writing any complex logic.

Installation

Install the package via pip. Requires Python 3.9+.

pip install freeflow-llm

Quick Start

# Linux/macOS
export GROQ_API_KEY='["gsk_key1...", "gsk_key2..."]'
export GEMINI_API_KEY='["AIzaSy...", "AIzaSy..."]'

# Windows (PowerShell)
$env:GROQ_API_KEY='["gsk_key1...", "gsk_key2..."]'
$env:GEMINI_API_KEY='["AIzaSy...", "AIzaSy..."]'

from freeflow_llm import FreeFlowClient

# Initialize client (automatically finds keys in env)
with FreeFlowClient() as client:
    response = client.chat(
        messages=[
            {"role": "user", "content": "Explain quantum computing in one sentence."}
        ]
    )
    
    print(f"AI: {response.content}")

Advanced Usage

with FreeFlowClient() as client:
    stream = client.chat_stream(
        messages=[{"role": "user", "content": "Write a story..."}]
    )
    
    for chunk in stream:
        print(chunk.content, end="", flush=True)

export GROQ_API_KEY='["key_1", "key_2", "key_3"]'

API Reference

Frequently Asked Questions