Google Gemini API Node.js Server

This Node.js server is designed to consume Google's Gemini API, a powerful generative AI tool. It includes two primary functionalities:

1. Text Generation from Text-Only Input

   • Model Name: gemini-pro

2. Text Generation from Text-and-Image Input (Multimodal)

   • Model Name: gemini-pro-vision

Prerequisites

Before running the project, make sure you have the following:

• Node.js version 18 and above installed.
• Gemini API keys generated from Google AI Studio.

Getting Started

1. Clone the Repository:

   git clone https://github.com/samson-shukla/google-gemini-ai.git
   cd google-gemini-ai

2. Install Dependencies:

   npm install

3. Configure Settings:

   • Create a .env file in the project root with the following environment variables:

   PORT=3000
   GEMINI_API_KEY=your_api_key

   • Adjust other settings in the server/config folder if needed.

4. Run the Server:

   npm run start

Usage and Making API Calls

• The server will be running at http://localhost:3000 by default.
• Access the API endpoints by making a POST request at http://localhost:3000/chat-with-gemini
• The request body must contain the following: - "modelType" (valid values are "text_only" and "text_and_image") - "prompt" as a string where you ask what you want to ask. - "imageParts" is an array of image URLs of the images you want to ask your prompt for.
• For "text_only" model, modelType and prompt is required
• For "text_and_image" model, modelType, prompt and imageParts is required

Request body example of "text_and_image" model

{
  "modelType": "text_only",
  "prompt": "What can you do for me?"
}

Request body example of "text_and_image" model

{
  "modelType": "text_and_image",
  "prompt": "What do you see in these images?",
  "imageParts": ["https://firstImage.jpg", "https://secondImage.png"]
}

Integration with Frontend

To implement the views, you can integrate any frontend library of your choice, such as ReactJs, Swelte, etc.

Feel free to enhance and customize the project to meet your specific needs. If you encounter any issues or have suggestions for improvements, please refer to the GitHub repository and open an issue.

Happy coding! 🚀