AI Multi-API Text Summarizer

A Python-based LLM application that summarizes text into Traditional Chinese bullet points using multiple AI providers, with automatic fallback for improved service robustness.

---

Overview

This project demonstrates lightweight LLM service orchestration and API integration in a practical summarization scenario.

It supports multiple providers:

• Hugging Face
• Google Gemini
• OpenRouter

If one provider fails due to quota limits, service unavailability, or response issues, the system automatically switches to another available provider.

---

Key Features

• Multi-provider LLM integration
• Automatic fallback across providers
• Traditional Chinese bullet-point summarization
• Environment-based API key management with .env
• Simple CLI workflow for local testing

---

Project Value

This project demonstrates practical experience in:

• integrating multiple LLM APIs
• handling provider fallback logic
• managing environment-based configuration
• building a simple but extensible application workflow

---

System Design Perspective

This project is not focused on model training.
Instead, it emphasizes application-layer AI engineering:

• integrating multiple external AI services
• improving robustness through fallback routing
• structuring provider-specific logic into reusable modules
• building a practical text-processing workflow

---

Project Structure

main.py
providers/
├── huggingface_provider.py
├── gemini_provider.py
└── openrouter_provider.py

---

Installation

pip install -r requirements.txt

---

Setup

Create a .env file:

GEMINI_API_KEY=your_key
HF_API_KEY=your_key
OPENROUTER_API_KEY=your_key

---

Usage

python main.py

Enter your text and the program will return a summarized result.

---

Example

Input:

Artificial intelligence is changing education and healthcare. Companies are focusing on efficiency, cost, and data security.

Output:

1. AI is transforming education and healthcare
2. Companies focus on improving efficiency
3. Cost reduction and data security are key priorities

---

Notes

• API keys are required to run this project
• Free-tier APIs may have rate limits or availability issues
• The fallback mechanism is designed to improve reliability when one provider is temporarily unavailable

---

Author

This project was built to demonstrate practical skills in LLM API integration, fallback handling, and lightweight AI application design.