ImaGen

ImaGen is a web application that generates images from text prompts using the Google Gemini AI model. It features a React frontend and an Express/Node.js backend.

Architecture

• Frontend: React (powered by Vite)
• Backend: Node.js with Express.js
• AI Model: Google Gemini API (gemini-2.0-flash)

Model Used and Pricing

This project uses the Google Gemini gemini-2.0-flash model for image generation.

As of the current Gemini API pricing structure, the pricing for using gemini-2.0-flash is heavily dependent on whether you use the Free Tier or the Pay-as-you-go Tier.

Free Tier (Rate Limits Apply)

You can use the Gemini API for free, subject to the following limits:

• RPM (Requests Per Minute): 15
• RPD (Requests Per Day): 1,500
• Tokens Per Minute: 1 million

Note: When using the free tier, your input and output data may be used by Google to improve their products.

Pay-as-you-go Tier

If you exceed free tier limits or disable data sharing, you will transition to Pay-as-you-go pricing:

• Prompts up to 128K tokens:
  • Input: $0.10 per 1 million tokens
  • Output: $0.40 per 1 million tokens
  • Images: $0.000131 per image
• Prompts longer than 128K tokens:
  • Input: $0.20 per 1 million tokens
  • Output: $0.80 per 1 million tokens
  • Images: $0.000262 per image

(Note: Pricing is subject to change. Always refer to the official Google Gemini API pricing page for the latest information).

Getting Started

Prerequisites

• Node.js installed on your machine.
• A Google Gemini API key.

Setup Backend

1. Navigate to the backend directory:

   cd imagen-backend

2. Install dependencies:

   npm install

3. Create a .env file in the imagen-backend directory and add your API key:

   GEMINI_API_KEY=your_api_key_here

4. Start the backend server:

   node server.js

   The backend will run on http://localhost:5000.

Setup Frontend

1. Open a new terminal and navigate to the frontend directory:

   cd imagen-frontend

2. Install dependencies:

   npm install

3. Start the development server:

   npm run dev

4. Open the displayed local URL (usually http://localhost:5173) in your browser.