AI Code Summarizer

A full-stack AI-powered application that takes in code from users, sends it to Google's Gemini API for summarization, and displays a human-readable explanation of the logic. It supports multiple languages and different Gemini models, with response metadata such as token counts and model latency.

---

Watch it live :

https://ai-code-summarizer.jack1337.serv00.net/

⚙️ Tech Stack

Frontend

• React (Vite)
• Tailwind CSS
• Fetch API
• Dark theme UI

Backend

• Node.js
• Express.js
• Zod (for validation)
• Google GenAI SDK (Gemini API)
• dotenv
• CORS
• Express Rate Limit (for safety)

---

🚀 Features

• 🧠 AI-powered code summarization
• ✍️ Supports multiple programming languages
• 📈 Model selection (Gemini 2.5 Pro / Flash / Flash Lite)
• 📊 Displays model metadata:
  • Model version
  • Token counts (prompt, candidate, total, thoughts)
  • Response time in seconds
• 🌓 Dark themed, responsive UI
• 🔁 Loading animation while waiting for AI
• ✅ Input validation using Zod
• ⚠️ Rate-limiting friendly error handling
• 🐛 Friendly error UI messages on failure

---

🧪 Local Setup

1. Clone the Repository

git clone [email protected]:aradhyacp/Ai-Code-Summarizer.git
cd Ai-Code-Summarizer

• Frontend runs on: http://localhost:5173
• Backend runs on: http://localhost:3000

---

📤 API Reference

POST /ai-summary

• Request Body:

{
  "language": "javascript",
  "codeSent": "function test() { return 123; }",
  "model": "gemini-2.5-flash-lite"
}

• Success Response:

{
  "response": "This function returns the number 123.",
  "modelVersion": "gemini-2.5-flash-lite",
  "promptTokenCount": 100,
  "candidatesTokenCount": 50,
  "totalTokenCount": 160,
  "thoughtsTokenCount": 10,
  "modelResponseTime": 1.48
}

• Validation Error Response:

{
  "error": "Invalid input",
  "issues": [
    { "message": "Language is required" },
    { "message": "Code is required" },
    { "message": "Model is required" }
  ]
}

---

🛡️ Error Handling

• Zod validation errors are caught and returned to frontend

• Invalid inputs or empty fields are blocked

• Rate-limiting errors show:

Too many requests. Please try again later. Rate limiter is turned on for safety purpose, limit is 10 requests every minute.

• Errors during fetch are shown as in-code comments like // Error message here

---

🧠 Tips for Usage

• Always select both Language and Model before submitting

• Code must not be empty

• Wait for the response spinner to finish before editing again

• Use Gemini Flash-lite for faster (but cheaper) summaries